LF-YOLO (Lighter and Faster YOLO) is used to detect defect of X-ray weld image.

Last update: Dec 13, 2022

Related tags

Deep Learning LF-YOLO

Overview

This project is based on ultralytics/yolov3.

LF-YOLO (Lighter and Faster YOLO) is used to detect defect of X-ray weld image.

Download

$ git clone https://github.com/lmomoy/LF-YOLO

Train

We provide multiple versions of LF-YOLO with different depths.

$ python train.py --data coco.yaml --cfg LF-YOLO.yaml      --weights '' --batch-size 1
                                         LF-YOLO-1.25.yaml                           1
                                         LF-YOLO-0.75.yaml                           1
                                         LF-YOLO-0.5.yaml                            1

Results

We test our model on public dataset MS COCO, and it also achieves competitive results.

Model	size ^(pixels)	mAP^val 0.5:0.95	mAP^{test 0.5:0.95}	mAP^val 0.5	params ^(M)	FLOPS ^{640 (B)}
YOLOv3-tiny	640	17.6	17.6	34.8	8.8	13.2
YOLOv3	640	43.3	43.3	63.0	61.9	156.3
YOLOv3-SPP	640	44.3	44.3	64.6	63.0	157.1
LF-YOLO	640	27.8	27.9	47.8	7.4	17.1

Table Notes (click to expand)

AP^test denotes COCO test-dev2017 server results, all other AP results denote val2017 accuracy.
AP values are for single-model single-scale unless otherwise noted. Reproduce mAP by python test.py --data coco.yaml --img 640 --conf 0.001 --iou 0.65
Speed_GPU averaged over 5000 COCO val2017 images using a GCP n1-standard-16 V100 instance, and includes FP16 inference, postprocessing and NMS. Reproduce speed by python test.py --data coco.yaml --img 640 --conf 0.25 --iou 0.45
All checkpoints are trained to 300 epochs with default settings and hyperparameters (no autoaugmentation).

Requirements

Python 3.8 or later with all requirements.txt dependencies installed, including torch>=1.7. To install run:

$ pip install -r requirements.txt

Inference

$ python detect.py --source data/images --weights LF-YOLO.pt --conf 0.25

LF-YOLO (Lighter and Faster YOLO) is used to detect defect of X-ray weld image.

Related tags

Overview

Download

Train

Results

Requirements

Inference

Owner

Unofficial Implementation of RobustSTL: A Robust Seasonal-Trend Decomposition Algorithm for Long Time Series (AAAI 2019)

CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped

A deep learning based semantic search platform that computes similarity scores between provided query and documents

EdiBERT is a generative model based on a bi-directional transformer, suited for image manipulation

Official implementation of "A Unified Objective for Novel Class Discovery", ICCV2021 (Oral)

SIR model parameter estimation using a novel algorithm for differentiated uniformization.

A Tensorflow implementation of BicycleGAN.

Learning recognition/segmentation models without end-to-end training. 40%-60% less GPU memory footprint. Same training time. Better performance.

Unofficial implementation of Perceiver IO: A General Architecture for Structured Inputs & Outputs

Official code for "Decoupling Zero-Shot Semantic Segmentation"

Hide screen when boss is approaching.

Modular Gaussian Processes

VGGFace2-HQ - A high resolution face dataset for face editing purpose

Libraries, tools and tasks created and used at DeepMind Robotics.

Official implementation of Protected Attribute Suppression System, ICCV 2021

RoboDesk A Multi-Task Reinforcement Learning Benchmark

"Graph Neural Controlled Differential Equations for Traffic Forecasting", AAAI 2022

Reference models and tools for Cloud TPUs.

This repository contains the code for "SBEVNet: End-to-End Deep Stereo Layout Estimation" paper by Divam Gupta, Wei Pu, Trenton Tabor, Jeff Schneider

Deep Reinforcement Learning with pytorch & visdom