Repository accompanying the "Sign Pose-based Transformer for Word-level Sign Language Recognition" paper

Related tags

Deep Learningtransformerslrlsa64sign-language-recognitionwlasl
Overview

Alt Text

by Matyáš Boháček and Marek Hrúz, University of West Bohemia
Should you have any questions or inquiries, feel free to contact us here.

PWC

Repository accompanying the Sign Pose-based Transformer for Word-level Sign Language Recognition paper, where we present a novel architecture for word-level sign language recognition based on the Transformer model. We designed our solution with low computational cost in mind, since we see egreat potential in the usage of such recognition system on hand-held devices. We introduce multiple original augmentation techniques tailored for the task of sign language recognition and propose a unique normalization scheme based on sign language linguistics.

Alt Text

Get Started

First, make sure to install all necessary dependencies using:

pip install -r requirements.txt

To train the model, simply specify the hyperparameters and run the following:

python -m train
  --experiment_name [str; name of the experiment to name the output logs and plots]
  
  --epochs [int; number of epochs]
  --lr [float; learning rate]
  
  --training_set_path [str; path to the csv file with training set's skeletal data]
  --validation_set_path [str; path to the csv file with validation set's skeletal data]
  --testing_set_path [str; path to the csv file with testing set's skeletal data]

If either the validation or testing sets' paths are left empty, these corresponding metrics will not be calculated. We also provide out-of-the box parameter to split the validation set as a desired split of the training set while preserving the label distribution for datasets without author-specified splits. These and many other specific hyperparameters with their descriptions can be found in the train.py file. All of them are provided a default value we found to be working well in our experiments.

Data

As SPOTER works on top of sequences of signers' skeletal data extracted from videos, we wanted to eliminate the computational demands of such annotation for each training run by pre-collecting this. For this reason and reproducibility, we are open-sourcing this data for WLASL100 and LSA64 datasets along with the repository. You can find the data here.

Alt Text

License

The code is published under the Apache License 2.0 which allows for both academic and commercial use if relevant License and copyright notice is included, our work is cited and all changes are stated.

The accompanying skeletal data of the WLASL and LSA64 datasets used for experiments are, however, shared under the Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license allowing only for non-commercial usage.

Citation

If you find our work relevant, build upon it or compare your approaches with it, please cite our work as stated below:

@InProceedings{Bohacek_2022_WACV,
    author    = {Boh\'a\v{c}ek, Maty\'a\v{s} and Hr\'uz, Marek},
    title     = {Sign Pose-Based Transformer for Word-Level Sign Language Recognition},
    booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops},
    month     = {January},
    year      = {2022},
    pages     = {182-191}
}
Comments
  • Pose based GRU model

    Pose based GRU model

    Thank you for providing this dataset. I'm trying to reproduce your results using the Pose based GRU model however I'm unable to do so. Could you please share the model architecture and hyperparameters. It would be quite helpful

    EDIT: Wrong repository, please delete this issue

    opened by farhaan-mukarram 0
  • Testing new data

    Testing new data

    I'm trying to use the model, but I'm having problems with the following step. I have already trained the model, and now I have the checkpoint_v_0.pth file, with which I do the following to load the generated model:

    model = torch.load(PATH/to/pth/file)
print(model)

    And this returns something like:

    SPOTER(
  (transformer): Transformer(
    (encoder): TransformerEncoder(
      (layers): ModuleList(
        (0): TransformerEncoderLayer(
          (self_attn): MultiheadAttention(
            (out_proj): _LinearWithBias(in_features=108, out_features=108, bias=True)
          )
          (linear1): Linear(in_features=108, out_features=2048, bias=True)
          (dropout): Dropout(p=0.1, inplace=False)
          (linear2): Linear(in_features=2048, out_features=108, bias=True)
          (norm1): LayerNorm((108,), eps=1e-05, elementwise_affine=True)
          (norm2): LayerNorm((108,), eps=1e-05, elementwise_affine=True)
          (dropout1): Dropout(p=0.1, inplace=False)
          (dropout2): Dropout(p=0.1, inplace=False)
        )
        (1): TransformerEncoderLayer(
...

    which makes me believe that everything is OK to this point.

    Now I would like to see how I can use the model to make a prediction with new data, and there's where the problem is.

    When running model(input) to get the results, some errors appear, and I believe that I'm not giving the correct kind of input. I'm using the second line from WLASL100_train_25fps.csv, changing the "s for [s in order to get a hierarchy like the following:

    [
           [a, b, c],
           [d],
           [e, f]
]

    However, this doesn't seem to work. Am I using a different format to the one the model should be given?

    The exact input I'm giving the model follows, with the np.array conversion:

    parsed_example = [[0.398577,0.398577,0.398577,0,0,0,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.398577,0.393627,0.398276,0.402398,0.400453,0.392099,0.389084,0.390115,0.390122,0.389504,0.389435,0.391493,0,0,0.404229,0,0.375582],[0.492837,0.492837,0.492837,0,0,0,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.492837,0.493243,0.493652,0.493095,0.493131,0.489589,0.487246,0.482847,0.461756,0.449853,0.454416,0.470932,0,0,0,0,0],[0.528982,0.528982,0.528982,0,0,0,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.528982,0.51803,0.497051,0.526573,0.514082,0.518499,0.509619,0.49415,0.472887,0.446121,0.418547,0.298164,0,0,0.0109521,0,0.132508],[0.471492,0.471492,0.471492,0,0,0,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.471492,0.47169,0.471791,0.472173,0.471784,0.471199,0.468871,0.464683,0.457271,0.451461,0.444686,0.442517,0,0,0,0,0],[0.0485368,0.0485368,0.0485368,0,0,0,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0485368,0.0484803,0.0485052,0.0437046,0.0458512,0.0410389,0.0284424,0.0229825,0.0172411,0.0192034,0.0174977,0.0778546,0,0,0,0,0],[0.444718,0.444718,0.444718,0,0,0,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.444718,0.410827,0.403959,0.461306,0.422898,0.468788,0.459742,0.435957,0.406939,0.3709,0.33533,0.238105,0,0,0,0,0.0313517],[0.595102,0.595102,0.595102,0,0,0,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.595102,0.555913,0.551131,0.624477,0.588528,0.650886,0.647908,0.627087,0.602364,0.573864,0.529823,0.44072,0,0,0.0573245,0,0.169048],[0.171264,0.171264,0.171264,0,0,0,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171264,0.171651,0.173042,0.170316,0.171098,0.165272,0.156861,0.150788,0.145757,0.128707,0.110712,0.070122,0,0,0,0,0],[0.473231,0.473231,0.473231,0,0,0,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.473231,0.45867,0.453409,0.493301,0.457943,0.478756,0.466767,0.4517,0.428147,0.40929,0.380941,0.318115,0,0,0.134512,0,0.16755],[0.536177,0.536177,0.536177,0,0,0,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.536177,0.517749,0.51461,0.529266,0.574933,0.532947,0.521931,0.502335,0.482224,0.459312,0.427157,0.32395,0,0,0.0160242,0,0.185826],[0.407958,0.407958,0.407958,0,0,0,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.407958,0.409048,0.409384,0.408326,0.408705,0.407319,0.405705,0.402739,0.395891,0.385984,0.3856,0.432976,0,0,0,0,0],[0.077801,0.077801,0.077801,0.0807571,0.0807571,0.0807571,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.077801,0.0762554,0.0732151,0.0731308,0.0731247,0.0724378,0.0797846,0.0793943,0.0781348,0.0832884,0.072022,0.0677893,0.0737613,0.0725894,0.0822043,0.0762998,0.0776332],[0.422698,0.422698,0.422698,0,0,0,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.422698,0.430718,0.440697,0.419992,0.431422,0.406726,0.40075,0.401792,0.405708,0.405919,0.404651,0.395599,0,0,0,0,0.402912],360,[0.556073,0.556073,0.556073,0,0,0,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.556073,0.536601,0.530958,0.55007,0.569923,0.549784,0.538729,0.52079,0.500619,0.477842,0.446324,0.330212,0,0,0.0192175,0,0.186272],[0.397988,0.397988,0.397988,0,0,0,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.397988,0.388859,0.386205,0.396229,0.380661,0.403792,0.401147,0.377418,0.346069,0.317356,0.289145,0.204902,0,0,0.0614293,0,0.119065],[0.417861,0.417861,0.417861,0.439912,0.439912,0.439912,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417861,0.417895,0.418117,0.418545,0.4183,0.419513,0.418768,0.420636,0.422949,0.424417,0.425064,0.427494,0.428411,0.428882,0.432829,0.438765,0.430258],[0.392557,0.392557,0.392557,0,0,0,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.392557,0.393167,0.393362,0.392457,0.393228,0.391265,0.388041,0.384542,0.381818,0.375773,0.375523,0.419973,0,0,0,0,0],[0.559241,0.559241,0.559241,0,0,0,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.559241,0.558091,0.557011,0.557938,0.557323,0.55621,0.556297,0.553147,0.548757,0.536598,0.533963,0.400428,0,0,0,0,0],[0.868213,0.868213,0.868213,0.866513,0.866513,0.866513,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868213,0.868191,0.868267,0.867593,0.868017,0.867536,0.865815,0.865587,0.866913,0.866851,0.866919,0.866125,0.864815,0.859497,0.851707,0.846945,0.855105],[0.498047,0.498047,0.498047,0.485291,0.485291,0.485291,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.498047,0.50619,0.497473,0.49691,0.496445,0.495318,0.495816,0.497309,0.500691,0.500798,0.495025,0.489368,0.486047,0.483157,0.477592,0.475032,0.480784],[0.0266556,0.0266556,0.0266556,0,0,0,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0266556,0.0267041,0.0261174,0.0241547,0.0247721,0.0226099,0.01849,0.0169849,0.0164648,0.020179,0.0184484,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0.418166,0.418166,0.418166,0,0,0,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.418166,0.426177,0.437865,0.416506,0.430483,0.401749,0.395495,0.397367,0.400526,0.402528,0.402191,0.416395,0,0,0.421347,0,0.364901],[0.510201,0.510201,0.510201,0,0,0,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.510201,0.483359,0.472029,0.47806,0.504377,0.473304,0.466987,0.44647,0.425754,0.398267,0.372791,0.26556,0,0,0,0,0.0290775],[0.393632,0.393632,0.393632,0,0,0,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.393632,0.394368,0.394109,0.393648,0.39417,0.392089,0.388192,0.385397,0.380724,0.376536,0.373512,0.510908,0,0,0,0,0],[0.514143,0.514143,0.514143,0,0,0,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.514143,0.488864,0.478717,0.499149,0.490987,0.487244,0.47975,0.463163,0.442731,0.411967,0.38342,0.262672,0,0,0,0,0.029516],[0.14742,0.14742,0.14742,0,0,0,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.14742,0.147526,0.148093,0.143568,0.145532,0.137696,0.122592,0.113845,0.095867,0.0645544,0.0335655,0.181972,0,0,0,0,0],[0.160598,0.160598,0.160598,0,0,0,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.160598,0.16084,0.161477,0.15815,0.159689,0.153363,0.14376,0.138518,0.120059,0.0925082,0.0678847,0.123008,0,0,0,0,0],[0.354376,0.354376,0.354376,0,0,0,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.354376,0.355446,0.357547,0.356205,0.355636,0.346679,0.342398,0.341331,0.341753,0.339144,0.337713,0.333239,0,0,0.416292,0,0.370757],[0.420521,0.420521,0.420521,0,0,0,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.420521,0.421522,0.421487,0.42075,0.421606,0.417578,0.413965,0.408728,0.403928,0.385775,0.397049,0,0,0,0,0,0],640,[0.541161,0.541161,0.541161,0,0,0,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.541161,0.54111,0.540695,0.541796,0.540932,0.53986,0.539377,0.533142,0.524508,0.516081,0.512644,0.484296,0,0,0,0,0],[0.556609,0.556609,0.556609,0,0,0,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.556609,0.540239,0.524339,0.561196,0.541111,0.561099,0.551605,0.534581,0.511972,0.485199,0.455068,0.340676,0,0,0.0149052,0,0.178714],[0.408994,0.408994,0.408994,0,0,0,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.408994,0.410046,0.409876,0.409208,0.409532,0.408416,0.40626,0.404551,0.396929,0.388693,0.380822,0.492037,0,0,0,0,0],[0.501293,0.501293,0.501293,0,0,0,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.501293,0.500814,0.500966,0.502335,0.501341,0.4978,0.496459,0.484296,0.466156,0.462022,0.462566,0.470642,0,0,0,0,0],[0.568233,0.568233,0.568233,0,0,0,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.568233,0.534232,0.531382,0.633149,0.593982,0.662144,0.663524,0.646581,0.624724,0.596565,0.546316,0.426962,0,0,0.0220607,0,0.160657],[0.466392,0.466392,0.466392,0,0,0,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.466392,0.426378,0.417419,0.498705,0.439079,0.504048,0.493154,0.475294,0.447406,0.417194,0.381258,0.290019,0,0,0,0,0.13123],[0.416737,0.416737,0.416737,0,0,0,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.416737,0.417708,0.41797,0.416587,0.417454,0.413279,0.416709,0.412278,0.402862,0.396735,0.399356,0.435406,0,0,0,0,0],[0.505452,0.505452,0.505452,0,0,0,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505452,0.505407,0.505082,0.505261,0.505529,0.504117,0.501017,0.496164,0.489083,0.479575,0.473823,0.465006,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0.415748,0.415748,0.415748,0,0,0,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.415748,0.413991,0.41782,0.425666,0.42046,0.419226,0.417096,0.416879,0.413618,0.409052,0.407485,0.401017,0,0,0,0,0.381397],[0.473815,0.473815,0.473815,0.473295,0.473295,0.473295,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.473815,0.477564,0.47261,0.473035,0.472369,0.471813,0.472967,0.473136,0.473406,0.473849,0.470469,0.470649,0.470665,0.468623,0.472882,0.471804,0.470694],[0.834083,0.834083,0.834083,0.850918,0.850918,0.850918,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.834083,0.833912,0.833459,0.832564,0.83294,0.832221,0.831697,0.831018,0.831089,0.831369,0.833802,0.842826,0.844824,0.847571,0.8431,0.837603,0.846742],[0.449698,0.449698,0.449698,0.470586,0.470586,0.470586,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449698,0.449782,0.449879,0.450069,0.450108,0.449954,0.449856,0.450556,0.451792,0.452467,0.453156,0.45535,0.457895,0.461187,0.464011,0.469523,0.462232],[0.335874,0.335874,0.335874,0.331081,0.331081,0.331081,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.335874,0.341357,0.338583,0.340809,0.337994,0.338086,0.344122,0.345942,0.338647,0.337904,0.341081,0.336325,0.335902,0.333218,0.330601,0.329757,0.333555],[0.59528,0.59528,0.59528,0,0,0,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.59528,0.567118,0.557465,0.582364,0.582105,0.569597,0.559301,0.545132,0.527121,0.498238,0.473415,0.380939,0,0,0.104355,0,0.181015],[0.496958,0.496958,0.496958,0,0,0,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.496958,0.497091,0.497496,0.498371,0.497851,0.493398,0.491022,0.482091,0.461508,0.455779,0.458801,0.467427,0,0,0,0,0],[0.361768,0.361768,0.361768,0,0,0,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.361768,0.359179,0.359225,0.370511,0.365155,0.36131,0.354589,0.358833,0.354196,0.356534,0.360185,0.41851,0.363539,0.320566,0.366688,0,0.3458],[0.684028,0.684028,0.684028,0.679071,0.679071,0.679071,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.684028,0.685253,0.687811,0.687514,0.685538,0.68235,0.681399,0.682819,0.678458,0.676241,0.67612,0.682778,0.684522,0.685962,0.686465,0.688187,0.685772],[0.0544637,0.0544637,0.0544637,0,0,0,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0544637,0.0550249,0.0553504,0.0499061,0.0534384,0.044793,0.0322963,0.0245405,0.0181975,0.0202103,0.019411,0.170797,0,0,0,0,0],[0.0891951,0.0891951,0.0891951,0,0,0,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0891951,0.0932565,0.0959771,0.0958182,0.0954115,0.0875102,0.0879518,0.0898014,0.0899711,0,0,0,0,0,0,0,0],[0.53426,0.53426,0.53426,0,0,0,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.53426,0.518446,0.504989,0.518253,0.532188,0.513162,0.504528,0.487462,0.468042,0.444239,0.417316,0.296324,0,0,0.00939417,0,0.181328],[0.446253,0.446253,0.446253,0,0,0,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.446253,0.451696,0.455806,0.438116,0.451011,0.425048,0.420168,0.421585,0.425422,0.424959,0.42439,0.417447,0,0,0,0,0.409357],[0.102753,0.102753,0.102753,0,0,0,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102753,0.102729,0.102828,0.0987104,0.100436,0.0935981,0.0806946,0.0702772,0.054906,0.0299774,0.0198809,0.12341,0,0,0,0,0],[0.396484,0.396484,0.396484,0,0,0,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.396484,0.401827,0.409132,0.424724,0.412457,0.41568,0.421796,0.429313,0.436448,0.442276,0.439053,0.460547,0,0,0.434043,0,0.344103],[0.110049,0.110049,0.110049,0,0,0,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.110049,0.109349,0.109123,0.105126,0.106634,0.0995601,0.0802612,0.0683592,0.0451567,0.0277143,0.0186325,0.111674,0,0,0,0,0],[0.539246,0.539246,0.539246,0.546963,0.546963,0.546963,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539246,0.539066,0.539134,0.539199,0.539123,0.539077,0.539273,0.53932,0.539677,0.540008,0.54017,0.540263,0.540521,0.540596,0.541036,0.541248,0.540771],[0.0548425,0.0548425,0.0548425,0,0,0,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0548425,0.0542418,0.0533832,0.0498548,0.051891,0.0463195,0.0273696,0.0221977,0.0176158,0.0217204,0.0170223,0,0,0,0,0,0],[0.542839,0.542839,0.542839,0,0,0,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.542839,0.524727,0.516811,0.553083,0.525625,0.537226,0.526861,0.512477,0.490337,0.47245,0.444906,0.377974,0,0,0.130359,0,0.18641],[0.150671,0.150671,0.150671,0,0,0,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150671,0.150852,0.150788,0.146919,0.14878,0.142768,0.126964,0.119384,0.102077,0.0744487,0.0384642,0.1673,0,0,0,0,0],[0.410095,0.410095,0.410095,0,0,0,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.410095,0.398424,0.393819,0.415969,0.402977,0.407646,0.398718,0.393197,0.368544,0.347764,0.330697,0.404279,0.230818,0.154058,0.0596943,0,0.100313],[0.416503,0.416503,0.416503,0,0,0,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.416503,0.422869,0.431879,0.413538,0.426203,0.400101,0.394413,0.395265,0.397295,0.397438,0.397063,0.396927,0,0,0.441229,0,0.382928],[0.43507,0.43507,0.43507,0,0,0,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.43507,0.438539,0.448614,0.428002,0.434548,0.414545,0.407394,0.411703,0.417309,0.418202,0.417991,0.422711,0,0,0.417827,0,0.370956],[0.821136,0.821136,0.821136,0.811291,0.811291,0.811291,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821136,0.821926,0.822495,0.822522,0.822497,0.823278,0.822569,0.82126,0.821443,0.821284,0.818062,0.813727,0.815449,0.810539,0.807021,0.798309,0.807851],[0.430851,0.430851,0.430851,0,0,0,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.430851,0.42683,0.433715,0.435988,0.434987,0.428295,0.424971,0.423909,0.422747,0.418467,0.415278,0.405242,0,0,0,0,0.405511],[0.101606,0.101606,0.101606,0,0,0,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101606,0.101141,0.101381,0.0968134,0.0981556,0.0916496,0.0752112,0.0632372,0.0494419,0.0262216,0.0183296,0.112508,0,0,0,0,0],[0.449582,0.449582,0.449582,0.461298,0.461298,0.461298,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.449582,0.448939,0.447747,0.44916,0.448292,0.448309,0.450119,0.448963,0.446121,0.4469,0.445914,0.451929,0.455284,0.454089,0.468171,0.468576,0.460603],[0.466496,0.466496,0.466496,0,0,0,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.466496,0.467094,0.467914,0.469095,0.470014,0.460429,0.457925,0.450468,0.43648,0.438498,0.433476,0.455836,0,0,0,0,0],[0.624708,0.624708,0.624708,0.622507,0.622507,0.622507,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.624708,0.623141,0.622646,0.62599,0.622981,0.622384,0.625291,0.624968,0.622767,0.621487,0.620373,0.619648,0.616427,0.614055,0.612757,0.612532,0.615279],[0.526177,0.526177,0.526177,0,0,0,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.526177,0.480287,0.470079,0.553857,0.489107,0.561176,0.552853,0.538499,0.507974,0.485248,0.449111,0.386123,0,0,0.107808,0,0.163175],[0.45481,0.45481,0.45481,0,0,0,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.45481,0.455389,0.455943,0.454923,0.45635,0.44954,0.447753,0.440691,0.423731,0.427265,0.423051,0.490448,0,0,0,0,0],[0.204234,0.204234,0.204234,0,0,0,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204234,0.204113,0.205368,0.202543,0.20299,0.196526,0.185895,0.178505,0.170746,0.148086,0.129567,0.0828415,0,0,0,0,0],[0.367898,0.367898,0.367898,0,0,0,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.367898,0.370057,0.374154,0.372622,0.372275,0.355544,0.350507,0.352565,0.354334,0.355713,0.353757,0.357839,0,0,0.428125,0,0.351858],[0.117691,0.117691,0.117691,0.138631,0.138631,0.138631,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.117691,0.121325,0.11973,0.115287,0.118254,0.111078,0.109886,0.109614,0.109523,0.1113,0.109874,0.108667,0.101558,0.0960157,0.130688,0.140281,0.108177],[0.778065,0.778065,0.778065,0.78798,0.78798,0.78798,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.778065,0.776055,0.774519,0.7694,0.774433,0.774804,0.77705,0.775125,0.774783,0.775015,0.775836,0.778349,0.781946,0.781645,0.777685,0.78165,0.779866],[0.51824,0.51824,0.51824,0,0,0,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.51824,0.501449,0.49275,0.500936,0.528518,0.500707,0.491616,0.472689,0.452038,0.428517,0.400894,0.287902,0,0,0.00840527,0,0.108004],[0.406896,0.406896,0.406896,0,0,0,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.406896,0.414483,0.423803,0.41058,0.423448,0.394538,0.388644,0.389319,0.391539,0.391864,0.391256,0.405444,0,0,0.435431,0,0.36059],[0.439912,0.439912,0.439912,0,0,0,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.439912,0.446466,0.451574,0.429526,0.441795,0.417601,0.412222,0.413791,0.417597,0.418591,0.418759,0.418723,0,0,0.412832,0,0.413656],[0.261884,0.261884,0.261884,0,0,0,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.261884,0.262196,0.263839,0.258682,0.260631,0.253516,0.240025,0.231006,0.215567,0.187305,0.160848,0.101928,0,0,0,0,0],[0.579374,0.579374,0.579374,0,0,0,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579374,0.579022,0.57576,0.573264,0.57638,0.579428,0.569769,0.553384,0.539439,0,0,0,0,0,0,0,0],[0.230018,0.230018,0.230018,0.265555,0.265555,0.265555,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.230018,0.228681,0.227422,0.228515,0.226399,0.23087,0.23231,0.22904,0.228337,0.22853,0.227974,0.225432,0.225165,0.223333,0.237067,0.253136,0.224881],[0.815262,0.815262,0.815262,0.821639,0.821639,0.821639,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815262,0.815063,0.814983,0.813652,0.814503,0.813854,0.811651,0.81111,0.811488,0.81148,0.812017,0.815096,0.815878,0.813429,0.808218,0.797769,0.812485],[0.425065,0.425065,0.425065,0,0,0,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.425065,0.433385,0.442779,0.419189,0.434403,0.406565,0.40141,0.402252,0.404794,0.406132,0.406414,0.407944,0,0,0.432907,0,0.398722],[0.112402,0.112402,0.112402,0,0,0,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112402,0.112145,0.113432,0.110913,0.112488,0.107522,0.103476,0.0970407,0.0860801,0.066747,0.0423101,0.223387,0,0,0,0,0],[0.430419,0.430419,0.430419,0,0,0,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.430419,0.442212,0.448338,0.426143,0.440351,0.413631,0.408111,0.408833,0.412923,0.413494,0.412733,0.40684,0,0,0,0,0.404998],[0.520902,0.520902,0.520902,0,0,0,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.520902,0.496065,0.482911,0.493291,0.507911,0.484245,0.477791,0.459485,0.441084,0.41259,0.385504,0.268533,0,0,0,0,0.0289197],[0.393807,0.393807,0.393807,0.402961,0.402961,0.402961,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.393807,0.395141,0.395114,0.396246,0.395592,0.397371,0.39706,0.398014,0.398622,0.398965,0.399404,0.400009,0.399706,0.39797,0.399346,0.39906,0.398453],[0.0241121,0.0241121,0.0241121,0,0,0,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0241121,0.0243541,0.0245614,0.0226802,0.0236107,0.018773,0.0148085,0.0147978,0.0150427,0.0182812,0.0208316,0.177875,0,0,0,0,0],[0.491392,0.491392,0.491392,0,0,0,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.491392,0.444631,0.434354,0.524604,0.459176,0.538472,0.526777,0.511007,0.477358,0.450819,0.412703,0.335655,0,0,0,0,0.156115],[0.429912,0.429912,0.429912,0,0,0,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.429912,0.427394,0.435999,0.420681,0.421897,0.406905,0.408883,0.411225,0.414191,0.414511,0.414761,0.421739,0,0,0.403948,0,0.352846],[0.0237761,0.0237761,0.0237761,0,0,0,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237761,0.0237831,0.0237747,0.0218054,0.0223853,0.0203356,0.01715,0.0162433,0.0157508,0.0177439,0.0189165,0.17213,0,0,0,0,0],[0.443515,0.443515,0.443515,0,0,0,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.443515,0.4443,0.445452,0.443732,0.444447,0.441558,0.438953,0.43511,0.423936,0.412579,0.408079,0.464436,0,0,0,0,0],[0.45634,0.45634,0.45634,0,0,0,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.45634,0.456923,0.457293,0.457284,0.45688,0.455834,0.452731,0.448578,0.443363,0.434433,0.424495,0.419894,0,0,0,0,0],[0.477422,0.477422,0.477422,0.501547,0.501547,0.501547,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477422,0.477432,0.47748,0.477331,0.477479,0.477313,0.477678,0.47875,0.480577,0.48196,0.482681,0.487939,0.489213,0.490749,0.495524,0.50114,0.492736],[0.0638131,0.0638131,0.0638131,0,0,0,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0638131,0.0636309,0.0632372,0.0616828,0.0624193,0.0568239,0.044646,0.0367491,0.0279695,0.0211512,0.0198449,0.152649,0,0,0,0,0],[0.411805,0.411805,0.411805,0,0,0,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.411805,0.416394,0.424926,0.409986,0.406284,0.390013,0.381802,0.383461,0.391434,0.393855,0.391449,0.400279,0,0,0.434856,0,0.346088],25,[0.368108,0.368108,0.368108,0,0,0,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.368108,0.363292,0.366387,0.364083,0.36808,0.356994,0.356186,0.353955,0.350064,0.344502,0.342,0.329341,0,0,0.355768,0,0.378391],[0.399114,0.399114,0.399114,0,0,0,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399114,0.399894,0.399905,0.399106,0.3997,0.397057,0.395637,0.391855,0.390962,0.380463,0.383624,0,0,0,0,0,0],[0.110802,0.110802,0.110802,0,0,0,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.110802,0.111332,0.111822,0.108947,0.11035,0.103296,0.0906004,0.0859786,0.062927,0.0391179,0.0310239,0.133727,0,0,0,0,0],[0.41466,0.41466,0.41466,0,0,0,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.41466,0.417034,0.423521,0.423729,0.418714,0.413116,0.416772,0.421493,0.426379,0.428085,0.428693,0.448324,0,0,0.429682,0,0.346237],[0.435688,0.435688,0.435688,0,0,0,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.435688,0.43697,0.437748,0.43703,0.438256,0.430639,0.428863,0.423764,0.415232,0.409347,0.412642,0.459792,0,0,0,0,0],[0.442893,0.442893,0.442893,0,0,0,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.442893,0.439508,0.444029,0.446504,0.446885,0.434358,0.430223,0.429864,0.430755,0.426764,0.423663,0.411071,0,0,0,0,0.415579],[0.587632,0.587632,0.587632,0,0,0,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.587632,0.562825,0.55653,0.573737,0.564234,0.566475,0.555575,0.538977,0.519672,0.495886,0.470314,0.389479,0,0,0.11266,0,0.18808],[0.480291,0.480291,0.480291,0.476794,0.476794,0.476794,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.480291,0.482249,0.480615,0.4834,0.480488,0.480235,0.484707,0.485455,0.480707,0.479695,0.480727,0.477987,0.476165,0.473637,0.471679,0.471144,0.474417],[0.487587,0.487587,0.487587,0,0,0,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.487587,0.488348,0.488563,0.487894,0.488274,0.485159,0.483058,0.480062,0.464431,0.450689,0.44785,0.465273,0,0,0,0,0],[0.441586,0.441586,0.441586,0,0,0,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.441586,0.442258,0.442371,0.441065,0.441773,0.439819,0.436891,0.435743,0.424649,0.412579,0.399066,0.476791,0,0,0,0,0],[0.0318525,0.0318525,0.0318525,0,0,0,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0318525,0.0316939,0.0309895,0.0295434,0.0300431,0.0269248,0.0199101,0.0173464,0.0180922,0.01764,0.0191318,0.196985,0,0,0,0,0],[0.388652,0.388652,0.388652,0,0,0,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.388652,0.395811,0.404267,0.393654,0.39265,0.373147,0.367536,0.370049,0.374525,0.375855,0.375904,0.382444,0,0,0.430766,0,0.346908],[0.58603,0.58603,0.58603,0,0,0,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.58603,0.552916,0.54602,0.603386,0.561927,0.626661,0.622056,0.599757,0.571055,0.543242,0.504178,0.433521,0,0,0.109474,0,0.172189],[0.227427,0.227427,0.227427,0,0,0,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.227427,0.226712,0.228167,0.224049,0.225356,0.218891,0.205752,0.195831,0.184292,0.158103,0.132525,0.0926074,0,0,0,0,0],61,[0.145215,0.145215,0.145215,0,0,0,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.145215,0.144662,0.145876,0.140813,0.142778,0.134423,0.11845,0.107384,0.0866141,0.0551191,0.0283827,0.190118,0,0,0,0,0]]

input = np.array([np.array(sublist) for sublist in parsed_example])
    opened by RodGal-2020 0
  • IndexError when training model

    IndexError when training model

    This is the command used to train the model : python -m train --experiment_name "Spoter" --training_set_path "data/WLASL100_train_25fps.csv" --validation_set_path "data/WLASL100_val_25fps.csv" --testing_set_path "data/WLASL100_test_25fps.csv"

    I get the following error after the program runs for awhile:

    Starting Spoter... Traceback (most recent call last): File "/usr/lib/python3.7/runpy.py", line 193, in _run_module_as_main "main", mod_spec) File "/usr/lib/python3.7/runpy.py", line 85, in _run_code exec(code, run_globals) File "/content/drive/MyDrive/Spoter/train.py", line 272, in train(args) File "/content/drive/MyDrive/Spoter/train.py", line 174, in train train_loss, _, _, train_acc = train_epoch(slrt_model, train_loader, cel_criterion, sgd_optimizer, device) File "/content/drive/MyDrive/Spoter/spoter/utils.py", line 19, in train_epoch loss = criterion(outputs[0], labels[0]) File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 1102, in _call_impl return forward_call(*input, **kwargs) File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/loss.py", line 1152, in forward label_smoothing=self.label_smoothing) File "/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py", line 2846, in cross_entropy return torch._C._nn.cross_entropy_loss(input, target, weight, _Reduction.get_enum(reduction), ignore_index, label_smoothing) IndexError: Target 78 is out of bounds.

    Changing parameters like the epochs and learning rate does not fix the issue.

    question 
    opened by adhithiyaa-git 3
  • Problematic normalization

    Problematic normalization

    Screen Shot 2022-02-13 at 5 36 52 PM Got a validation accuracy around 58%, lower than the one proposed in the paper. Is the lower accuracy caused by this problematic normalization error?

    bug 
    opened by Coco-hanqi 6
  • Thank for your work! Please comment,when training ,report another error.

    Thank for your work! Please comment,when training ,report another error.

    RuntimeError: CUDA error: device-side assert triggered. ` for i, data in enumerate(dataloader): inputs, labels = data # inputs, labels = Variable(inputs), Variable(labels)-1 inputs = inputs.squeeze(0).to(device) labels = labels.to(device, dtype=torch.long)

        optimizer.zero_grad()
    outputs = model(inputs).expand(1, -1, -1)

    loss = criterion(outputs[0], labels[0])`
    bug 
    opened by showfaker66 5
Releases(supplementary-data)

  • supplementary-data(Dec 9, 2021)

    As SPOTER works on top of sequences of signers' skeletal data extracted from videos, we wanted to eliminate the computational demands of such annotation for each training run by pre-collecting this. For this reason and reproducibility, we are open-sourcing this data along with the code as well.

    This data is shared under the Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license allowing only for non-commercial usage only.

    We employed the WLASL100 and LSA64 datasets for our experiments. Their corresponding citations can be found below:

    @inproceedings{li2020word,
    title={Word-level Deep Sign Language Recognition from Video: A New Large-scale Dataset and Methods Comparison},
    author={Li, Dongxu and Rodriguez, Cristian and Yu, Xin and Li, Hongdong},
    booktitle={The IEEE Winter Conference on Applications of Computer Vision},
    pages={1459--1469},
    year={2020}
}

    @inproceedings{ronchetti2016lsa64,
    title={LSA64: an Argentinian sign language dataset},
    author={Ronchetti, Franco and Quiroga, Facundo and Estrebou, C{\'e}sar Armando and Lanzarini, Laura Cristina and Rosete, Alejandro},
    booktitle={XXII Congreso Argentino de Ciencias de la Computaci{\'o}n (CACIC 2016).},
    year={2016}
}
    Source code(tar.gz)
    Source code(zip)
    LSA64_60fps.csv(185.14 MB)
    WLASL100_test_25fps.csv(10.37 MB)
    WLASL100_train_25fps.csv(57.16 MB)
    WLASL100_val_25fps.csv(13.57 MB)
Owner
Matyáš Boháček
ML&NLP Researcher at @dataclair • Research Fellow with the University of West Bohemia •  WWDC19 & 21 Scholarship Winner
Matyáš Boháček
GitHub Repository https://spoter.signlanguagerecognition.com
Summary Explorer is a tool to visually explore the state-of-the-art in text summarization.

Summary Explorer Summary Explorer is a tool to visually inspect the summaries from several state-of-the-art neural summarization models across multipl

Webis 42 Aug 14, 2022
Attention-based Transformation from Latent Features to Point Clouds (AAAI 2022)

Attention-based Transformation from Latent Features to Point Clouds This repository contains a PyTorch implementation of the paper: Attention-based Tr

12 Nov 11, 2022
“袋鼯麻麻——智能购物平台”能够精准地定位识别每一个商品

“袋鼯麻麻——智能购物平台”能够精准地定位识别每一个商品，并且能够返回完整地购物清单及顾客应付的实际商品总价格，极大地降低零售行业实际运营过程中巨大的人力成本，提升零售行业无人化、自动化、智能化水平。

thomas-yanxin 192 Jan 05, 2023
Log4j JNDI inj. vuln scanner

Log-4-JAM - Log 4 Just Another Mess Log4j JNDI inj. vuln scanner Requirements pip3 install requests_toolbelt Usage # make sure target list has http/ht

Ashish Kunwar 66 Nov 09, 2022
Bagua is a flexible and performant distributed training algorithm development framework.

Bagua is a flexible and performant distributed training algorithm development framework.

786 Dec 17, 2022
ResNEsts and DenseNEsts: Block-based DNN Models with Improved Representation Guarantees

ResNEsts and DenseNEsts: Block-based DNN Models with Improved Representation Guarantees This repository is the official implementation of the empirica

Kuan-Lin (Jason) Chen 2 Oct 02, 2022
The implementation for "Comprehensive Knowledge Distillation with Causal Intervention".

Comprehensive Knowledge Distillation with Causal Intervention This repository is a PyTorch implementation of "Comprehensive Knowledge Distillation wit

Xiang Deng 10 Nov 03, 2022
Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow

Mask R-CNN for Object Detection and Segmentation This is an implementation of Mask R-CNN on Python 3, Keras, and TensorFlow. The model generates bound

Matterport, Inc 22.5k Jan 04, 2023
A project for developing transformer-based models for clinical relation extraction

Clinical Relation Extration with Transformers Aim This package is developed for researchers easily to use state-of-the-art transformers models for ext

uf-hobi-informatics-lab 101 Dec 19, 2022
Decorators for maximizing memory utilization with PyTorch & CUDA

torch-max-mem This package provides decorators for memory utilization maximization with PyTorch and CUDA by starting with a maximum parameter size and

Max Berrendorf 10 May 02, 2022
Machine Learning From Scratch. Bare bones NumPy implementations of machine learning models and algorithms with a focus on accessibility. Aims to cover everything from linear regression to deep learning.

Machine Learning From Scratch About Python implementations of some of the fundamental Machine Learning models and algorithms from scratch. The purpose

Erik Linder-Norén 21.8k Jan 09, 2023
Bayesian algorithm execution (BAX)

Bayesian Algorithm Execution (BAX) Code for the paper: Bayesian Algorithm Execution: Estimating Computable Properties of Black-box Functions Using Mut

Willie Neiswanger 38 Dec 08, 2022
Face and Pose detector that emits MQTT events when a face or human body is detected and not detected.

Face Detect MQTT Face or Pose detector that emits MQTT events when a face or human body is detected and not detected. I built this as an alternative t

Jacob Morris 38 Oct 21, 2022
ECCV2020 paper: Fashion Captioning: Towards Generating Accurate Descriptions with Semantic Rewards. Code and Data.

This repo contains some of the codes for the following paper Fashion Captioning: Towards Generating Accurate Descriptions with Semantic Rewards. Code

Xuewen Yang 56 Dec 08, 2022
Hierarchical Memory Matching Network for Video Object Segmentation (ICCV 2021)

Hierarchical Memory Matching Network for Video Object Segmentation Hongje Seong, Seoung Wug Oh, Joon-Young Lee, Seongwon Lee, Suhyeon Lee, Euntai Kim

Hongje Seong 72 Dec 14, 2022
Scale-aware Automatic Augmentation for Object Detection (CVPR 2021)

SA-AutoAug Scale-aware Automatic Augmentation for Object Detection Yukang Chen, Yanwei Li, Tao Kong, Lu Qi, Ruihang Chu, Lei Li, Jiaya Jia [Paper] [Bi

DV Lab 182 Dec 29, 2022
Convolutional Neural Network for Text Classification in Tensorflow

This code belongs to the "Implementing a CNN for Text Classification in Tensorflow" blog post. It is slightly simplified implementation of Kim's Convo

Denny Britz 5.5k Jan 02, 2023
[CVPR'21] Learning to Recommend Frame for Interactive Video Object Segmentation in the Wild

IVOS-W Paper Learning to Recommend Frame for Interactive Video Object Segmentation in the Wild Zhaoyun Yin, Jia Zheng, Weixin Luo, Shenhan Qian, Hanli

SVIP Lab 38 Dec 12, 2022
Image Completion with Deep Learning in TensorFlow

Image Completion with Deep Learning in TensorFlow See my blog post for more details and usage instructions. This repository implements Raymond Yeh and

Brandon Amos 1.3k Dec 23, 2022
[IROS2021] NYU-VPR: Long-Term Visual Place Recognition Benchmark with View Direction and Data Anonymization Influences

NYU-VPR This repository provides the experiment code for the paper Long-Term Visual Place Recognition Benchmark with View Direction and Data Anonymiza

Automation and Intelligence for Civil Engineering (AI4CE) Lab @ NYU 22 Sep 28, 2022