Official Pytorch implementation of Online Continual Learning on Class Incremental Blurry Task Configuration with Anytime Inference (ICLR 2022)

Related tags

Deep Learningi-Blurry
Overview

The Official Implementation of CLIB (Continual Learning for i-Blurry)

Online Continual Learning on Class Incremental Blurry Task Configuration with Anytime Inference
Hyunseo Koh*, Dahyun Kim*, Jung-Woo Ha, Jonghyun Choi
ICLR 2022 [Paper]
(* indicates equal contribution)

Overview

Abstract

Despite rapid advances in continual learning, a large body of research is devoted to improving performance in the existing setups. While a handful of work do propose new continual learning setups, they still lack practicality in certain aspects. For better practicality, we first propose a novel continual learning setup that is online, task-free, class-incremental, of blurry task boundaries and subject to inference queries at any moment. We additionally propose a new metric to better measure the performance of the continual learning methods subject to inference queries at any moment. To address the challenging setup and evaluation protocol, we propose an effective method that employs a new memory management scheme and novel learning techniques. Our empirical validation demonstrates that the proposed method outperforms prior arts by large margins.

Results

Results of CL methods on various datasets, for online continual learning on i-Blurry-50-10 split, measured by metric. For more details, please refer to our paper.

Methods CIFAR10 CIFAR100 TinyImageNet ImageNet
EWC++ 57.34±2.10 35.35±1.96 22.26±1.15 24.81
BiC 58.38±0.54 33.51±3.04 22.80±0.94 27.41
ER-MIR 57.28±2.43 35.35±1.41 22.10±1.14 20.48
GDumb 53.20±1.93 32.84±0.45 18.17±0.19 14.41
RM 23.00±1.43 8.63±0.19 5.74±0.30 6.22
Baseline-ER 57.46±2.25 35.61±2.08 22.45±1.15 25.16
CLIB 70.26±1.28 46.67±0.79 23.87±0.68 28.16

Getting Started

To set up the environment for running the code, you can either use the docker container, or manually install the requirements in a virtual environment.

Using Docker Container (Recommended)

We provide the Docker image khs8157/iblurry on Docker Hub for reproducing the results. To download the docker image, run the following command:

docker pull khs8157/iblurry:latest

After pulling the image, you may run the container via following command:

docker run --gpus all -it --shm-size=64gb -v /PATH/TO/CODE:/PATH/TO/CODE --name=CONTAINER_NAME khs8157/iblurry:latest bash

Replace the arguments written in italic with your own arguments.

Requirements

  • Python3
  • Pytorch (>=1.9)
  • torchvision (>=0.10)
  • numpy
  • pillow~=6.2.1
  • torch_optimizer
  • randaugment
  • easydict
  • pandas~=1.1.3

If not using Docker container, install the requirements using the following command

pip install -r requirements.txt

Running Experiments

Downloading the Datasets

CIFAR10, CIFAR100, and TinyImageNet can be downloaded by running the corresponding scripts in the dataset/ directory. ImageNet dataset can be downloaded from Kaggle.

Experiments Using Shell Script

Experiments for the implemented methods can be run by executing the shell scripts provided in scripts/ directory. For example, you may run CL experiments using CLIB method by

bash scripts/clib.sh

You may change various arguments for different experiments.

  • NOTE: Short description of the experiment. Experiment result and log will be saved at results/DATASET/NOTE.
    • WARNING: logs/results with the same dataset and note will be overwritten!
  • MODE: CL method to be applied. Methods implemented in this version are: [clib, er, ewc++, bic, mir, gdumb, rm]
  • DATASET: Dataset to use in experiment. Supported datasets are: [cifar10, cifar100, tinyimagenet, imagenet]
  • N_TASKS: Number of tasks. Note that corresponding json file should exist in collections/ directory.
  • N: Percentage of disjoint classes in i-blurry split. N=100 for full disjoint, N=0 for full blurry. Note that corresponding json file should exist in collections/ directory.
  • M: Blurry ratio of blurry classes in i-blurry split. Note that corresponding json file should exist in collections/ directory.
  • GPU_TRANSFORM: Perform AutoAug on GPU, for faster running.
  • USE_AMP: Use automatic mixed precision (amp), for faster running and reducing memory cost.
  • MEM_SIZE: Maximum number of samples in the episodic memory.
  • ONLINE_ITER: Number of model updates per sample.
  • EVAL_PERIOD: Period of evaluation queries, for calculating .

Citation

If you used our code or i-blurry setup, please cite our paper.

@inproceedings{koh2022online,
  title={Online Continual Learning on Class Incremental Blurry Task Configuration with Anytime Inference},
  author={Koh, Hyunseo and Kim, Dahyun and Ha, Jung-Woo and Choi, Jonghyun},
  booktitle={ICLR},
  year={2022}
}

License

Copyright (C) 2022-present NAVER Corp.

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program.  If not, see <https://www.gnu.org/licenses/>.
Owner
NAVER AI
Official account of NAVER CLOVA AI Lab, Korea No.1 Industrial AI Research Group
NAVER AI
A repository that finds a person who looks like you by using face recognition technology.

Find Your Twin Hello everyone, I've always wondered how casting agencies do the casting for a scene where a certain actor is young or old for a movie

Cengizhan Yurdakul 3 Jan 29, 2022
Continuum Learning with GEM: Gradient Episodic Memory

Gradient Episodic Memory for Continual Learning Source code for the paper: @inproceedings{GradientEpisodicMemory, title={Gradient Episodic Memory

Facebook Research 360 Dec 27, 2022
Source code for 2021 ICCV paper "In-the-Wild Single Camera 3D Reconstruction Through Moving Water Surfaces"

In-the-Wild Single Camera 3D Reconstruction Through Moving Water Surfaces This is the PyTorch implementation for 2021 ICCV paper "In-the-Wild Single C

27 Dec 06, 2022
This is 2nd term discrete maths project done by UCU students that uses backtracking to solve various problems.

Backtracking Project Sponsors This is a project made by UCU students: Olha Liuba - crossword solver implementation Hanna Yershova - sudoku solver impl

Dasha 4 Oct 17, 2021
TensorFlow2 Classification Model Zoo playing with TensorFlow2 on the CIFAR-10 dataset.

Training CIFAR-10 with TensorFlow2(TF2) TensorFlow2 Classification Model Zoo. I'm playing with TensorFlow2 on the CIFAR-10 dataset. Architectures LeNe

Chia-Hung Yuan 16 Sep 27, 2022
A python program to hack instagram

hackinsta a program to hack instagram Yokoback_(instahack) is the file to open, you need libraries write on import. You run that file in the same fold

2 Jan 22, 2022
Implementation of CVAE. Trained CVAE on faces from UTKFace Dataset to produce synthetic faces with a given degree of happiness/smileyness.

Conditional Smiles! (SmileCVAE) About Implementation of AE, VAE and CVAE. Trained CVAE on faces from UTKFace Dataset. Using an encoding of the Smile-s

Raúl Ortega 3 Jan 09, 2022
An open source Jetson Nano baseboard and tools to design your own.

My Jetson Nano Baseboard This basic baseboard gives the user the foundation and the flexibility to design their own baseboard for the Jetson Nano. It

NVIDIA AI IOT 57 Dec 29, 2022
CarND-LaneLines-P1 - Lane Finding Project for Self-Driving Car ND

Finding Lane Lines on the Road Overview When we drive, we use our eyes to decide where to go. The lines on the road that show us where the lanes are a

Udacity 769 Dec 27, 2022
A repository built on the Flow software package to explore cyber-security attacks on intelligent transportation systems.

A repository built on the Flow software package to explore cyber-security attacks on intelligent transportation systems.

George Gunter 4 Nov 14, 2022
[ICLR 2022] Pretraining Text Encoders with Adversarial Mixture of Training Signal Generators

AMOS This repository contains the scripts for fine-tuning AMOS pretrained models on GLUE and SQuAD 2.0 benchmarks. Paper: Pretraining Text Encoders wi

Microsoft 22 Sep 15, 2022
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator

ONNX Runtime is a cross-platform inference and training machine-learning accelerator. ONNX Runtime inference can enable faster customer experiences an

Microsoft 8k Jan 04, 2023
This program will stylize your photos with fast neural style transfer.

Neural Style Transfer (NST) Using TensorFlow Demo TensorFlow TensorFlow is an end-to-end open source platform for machine learning. It has a comprehen

Ismail Boularbah 1 Aug 08, 2022
Repo 4 basic seminar §How to make human machine readable"

WORK IN PROGRESS... Notebooks from the Seminar: Human Machine Readable WS21/22 Introduction into programming Georg Trogemann, Christian Heck, Mattis

experimental-informatics 3 May 29, 2022
The Few-Shot Bot: Prompt-Based Learning for Dialogue Systems

Few-Shot Bot: Prompt-Based Learning for Dialogue Systems This repository includes the dataset, experiments results, and code for the paper: Few-Shot B

Andrea Madotto 103 Dec 28, 2022
Software Platform for solving and manipulating multiparametric programs in Python

PPOPT Python Parametric OPtimization Toolbox (PPOPT) is a software platform for solving and manipulating multiparametric programs in Python. This pack

10 Sep 13, 2022
[ICLR 2021] HW-NAS-Bench: Hardware-Aware Neural Architecture Search Benchmark

HW-NAS-Bench: Hardware-Aware Neural Architecture Search Benchmark Accepted as a spotlight paper at ICLR 2021. Table of content File structure Prerequi

72 Jan 03, 2023
Code for the paper "SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness" (NeurIPS 2021)

SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness (NeurIPS2021) This repository contains code for the paper "Smo

Jongheon Jeong 17 Dec 27, 2022
code for Grapadora research paper experimentation

Road feature embedding selection method Code for research paper experimentation Abstract Traffic forecasting models rely on data that needs to be sens

Eric López Manibardo 0 May 26, 2022
[IJCAI'21] Deep Automatic Natural Image Matting

Deep Automatic Natural Image Matting [IJCAI-21] This is the official repository of the paper Deep Automatic Natural Image Matting. Introduction | Netw

Jizhizi_Li 316 Jan 06, 2023