sannapersson / yolov3-pytorch

YOLOv3 in PyTorch

A quite minimal implementation of YOLOv3 in PyTorch spanning only around 800 lines of code (not including plot functions etc.) with support for training and evaluation and complete with helper functions for inference. There is currently pretrained weights for Pascal-VOC with MS COCO coming up. With minimal changes in the model with regards to the output format the original weights can also be loaded seamlessly.

Installation

Clone and install requirements

$ git clone https://github.com/SannaPersson/YOLOv3-PyTorch.git
$ cd YOLOv3-PyTorch
$ pip install requirements.txt

Download pretrained weights on Pascal-VOC

Pretrained weights for Pascal-VOC can be downloaded from this page: https://www.kaggle.com/sannapersson/yolov3-weights-for-pascal-voc-with-781-map

Dowload original weights

Download YOLOv3 weights from https://pjreddie.com/media/files/yolov3.weights. Save the weights to PyTorch format by running the model_with_weights.py file. Change line in train.py to import model_with_weights.py instead of model.py since the original output format is slightly different.

Download Pascal VOC dataset

Download the preprocessed dataset from link. Just unzip this in the main directory. The file structure of the dataset is a folder with images, a folder with corresponding text files containing the bounding boxes and class targets for each image and two csv-files containing the subsets of the data used for training and testing.

Training

Edit the config.py file to match the setup you want to use. Then run train.py

Results

Model	mAP @ 50 IoU
YOLOv3 (Pascal VOC)	78.2
YOLOv3 (MS-COCO)	Not done yet

The model was evaluated with confidence 0.2 and IOU threshold 0.45 using NMS.

YOLOv3 paper

The implementation is based on the following paper:

An Incremental Improvement

by Joseph Redmon, Ali Farhadi

Abstract

We present some updates to YOLO! We made a bunch of little design changes to make it better. We also trained this new network that’s pretty swell. It’s a little bigger than last time but more accurate. It’s still fast though, don’t worry. At 320 × 320 YOLOv3 runs in 22 ms at 28.2 mAP, as accurate as SSD but three times faster. When we look at the old .5 IOU mAP detection metric YOLOv3 is quite good. It achieves 57.9 AP50 in 51 ms on a Titan X, compared to 57.5 AP50 in 198 ms by RetinaNet, similar performance but 3.8× faster. As always, all the code is online at https://pjreddie.com/yolo/.

@article{yolov3,
  title={YOLOv3: An Incremental Improvement},
  author={Redmon, Joseph and Farhadi, Ali},
  journal = {arXiv},
  year={2018}
}

yolov3-pytorch's Issues

adjoin grid cell loss compute

According to yolo detection thought which cell the midpoint(center_x, center_y) falls in is responsible for detect the object, but in upper code not consider the adjoin grid cell, if they also have the greater than ignore_iou_thresh, the adjoin grid cell will also compute the loss. Because the code do not set their targets[scale_idx][anchor_on_scale, i, j, 0] = -1? I am looking forward to your answer. Thank you in advance.