YOLOV and YOLOV++ for video object detection.

Update

July. 30th, 2024: The pre-print version of the YOLOV++ paper is now available on Arxiv.
May. 8th, 2024: We release code, log and weights for YOLOV++.
April. 21th, 2024: Our enhanced model now achieves a 92.9 AP50(w.o post-processing) on the ImageNet VID dataset, thanks to a more robust backbone and algorithm improvements. It maintains a processing time of 26.5ms per image during batch inference on a 3090 GPU. Code release is forthcoming.

Introduction

YOLOV series are high performance video object detector. Please refer to YOLOV and YOLOV++ on Arxiv for more details.

This repo is an implementation of PyTorch version YOLOV and YOLOV++ based on YOLOX.

YOLOX Pretain Models on ImageNet VID

Model	size	mAP@50^val	Speed 2080Ti(batch size=1) (ms)	Speed 3090(batch size=32) (ms)	weights
YOLOX-s	576	69.5	9.4	1.4	google
YOLOX-l	576	76.1	14.8	4.2	google
YOLOX-x	576	77.8	20.4	-	google
YOLOX-SwinTiny	576	79.2	19.0	5.5	google
YOLOX-SwinBase	576	86.5	24.9	11.8	google
YOLOX-FocalLarge	576	89.7	42.2	25.7	-

Main result in YOLOV++

Model	size	mAP@50^val	Speed 3090(batch size=32) (ms)	weights	logs
YOLOV++ s	576	78.7	5.3	google	link
YOLOV++ l	576	84.2	7.6	google	-
YOLOV++ SwinTiny	576	85.6	8.4	google	link
YOLOV++ SwinBase	576	90.7	15.9	google	link
YOLOV++ FocalLarge	576	92.9	27.6	google	link
YOLOV++ FocalLarge + Post	576	93.2	-	-

Main result in YOLOV

Model	size	mAP@50^val	Speed 2080Ti(batch size=1) (ms)	weights
YOLOV-s	576	77.3	11.3	google
YOLOV-l	576	83.6	16.4	google
YOLOV-x	576	85.5	22.7	google
YOLOV-x + post	576	87.5	-	-

TODO

[x] Finish Swin-Transformer based experiments.
[ ] Release updated code, model and log.

Quick Start

Installation

Install YOLOV from source. ```shell git clone git@github.com:YuHengsss/YOLOV.git cd YOLOV ``` Create conda env. ```shell conda create -n yolov python=3.7 conda activate yolov pip install -r requirements.txt pip3 install -v -e . ```

Demo

Step1. Download a pretrained weights. Step2. Run yolov demos. For example: ```shell python tools/vid_demo.py -f [path to your yolov exp files] -c [path to your yolov weights] --path /path/to/your/video --conf 0.25 --nms 0.5 --tsize 576 --save_result ``` For online mode, exampled with yolov_l, you can run: ```shell python tools/yolov_demo_online.py -f ./exp/yolov/yolov_l_online.py -c [path to your weights] --path /path/to/your/video --conf 0.25 --nms 0.5 --tsize 576 --save_result ``` For yolox models, please use python tools/demo.py for inferencing.

Reproduce our results on VID

Step1. Download datasets and weights: Download ILSVRC2015 DET and ILSVRC2015 VID dataset from [IMAGENET](https://image-net.org/challenges/LSVRC/2015/2015-downloads) and organise them as follows: ```shell path to your datasets/ILSVRC2015/ path to your datasets/ILSVRC/ ``` Download our COCO-style annotations for [training](https://drive.google.com/file/d/1HhE4OAcc--CpjUj69JCRXzMvIRsR4ymM/view?usp=sharing), FGFA version training [annotation](https://drive.google.com/file/d/12ceMTsmwkCMCdjYSM268qYfQTQcCDYFU/view?usp=drive_link) and [video sequences](https://drive.google.com/file/d/1vJs8rLl_2oZOWCMJtk3a9ZJmdNn8cu-G/view?usp=sharing). Then, put them in these two directories: ```shell YOLOV/annotations/vid_train_coco.json YOLOV/annotations/ILSVRC_FGFA_COCO.json YOLOV/yolox/data/dataset/train_seq.npy ``` Change the data_dir in exp files to [path to your datasets] and Download our weights. Step2. Generate predictions and convert them to IMDB style for evaluation. ```shell python tools/val_to_imdb.py -f exps/yolov/yolov_x.py -c path to your weights/yolov_x.pth --fp16 --output_dir ./yolov_x.pkl ``` Evaluation process: ```shell python tools/REPPM.py --repp_cfg ./tools/yolo_repp_cfg.json --predictions_file ./yolov_x.pkl --evaluate --annotations_filename ./annotations/annotations_val_ILSVRC.txt --path_dataset [path to your dataset] --store_imdb --store_coco (--post) ``` (--post) indicates involving post-processing method. Then you will get: ```shell {'mAP_total': 0.8758871720817065, 'mAP_slow': 0.9059275666099181, 'mAP_medium': 0.8691557352372217, 'mAP_fast': 0.7459511040452989} ``` **Training example** ```shell python tools/vid_train.py -f exps/yolov/yolov_s.py -c weights/yoloxs_vid.pth --fp16 ``` **Roughly testing** ```shell python tools/vid_eval.py -f exps/yolov/yolov_s.py -c weights/yolov_s.pth --tnum 500 --fp16 ``` tnum indicates testing sequence number.

Annotation format

Details

**Training base detector** The train_coco.json is a COCO format annotation file. When trainig the base detector on your own dataset, try to convert the annotation to COCO format. **Training YOLOV Series** The train_seq.npy and val_seq.npy files are numpy arrays of lists. They can be loaded using the following command: ```shell numpy.load('./yolox/data/datasets/train_seq.npy',allow_pickle=True) ``` Each list contains the paths to all images in a video. The specific annotations(xml annotation in VID dataset) are loaded via these image paths, refer to https://github.com/YuHengsss/YOLOV/blob/f5a57ddea2f3660875d6d75fc5fa2ddbb95028a7/yolox/data/datasets/vid.py#L125 for more details.

Training on Custom Datasets

Details

1. Finetuing the base detector(YOLOX) on your custom dataset with COCO format annotation. You need to modify the YOLOX experiment file. For instance, the experiment file for the Imagenet VID dataset is modified as [this example](https://github.com/YuHengsss/YOLOV/blob/master/exps/swin_base/swin_tiny_vid.py). Initialized weights with COCO pretraining is essential for the performance, you can find these coco pretrained weights in YOLOX official repo (YOLOX-S~YOLOX-X) and this [huggingface repo](https://huggingface.co/YuhengSSS/YOLOV/tree/main) (YOLOX-SwinTiny and SwinBase). 2. Construct your dataset in the COCO format. Here is a template for the dataset structure (sourced from [OVIS](https://songbai.site/ovis/)): ```shell { "info" : info, "videos" : [video], "annotations" : [annotation] or None, "categories" : [category], } video{ "id" : int, "width" : int, "height" : int, "length" : int, "file_names" : [file_name], } annotation{ "id" : int, "video_id" : int, "category_id" : int, "areas" : [float or None], "bboxes" : [[x,y,width,height] or None], "iscrowd" : 0 or 1, } category{ "id" : int, "name" : str, "supercategory" : str, } ``` After preparing the COCO format dataset, we provide [code](https://github.com/YuHengsss/YOLOV/blob/8873e06cac9912c60c31ca2ef3061d0bfe5b2f36/yolox/data/datasets/ovis.py#L238) which converts the COCO format annotation for video object detection. You can construct your experiment file for YOLOV such as [YOLOVs_OVIS](https://github.com/YuHengsss/YOLOV/blob/master/exps/yolov_ovis/yolovs_ovis_75_75_750.py). For YOLOV++, please refer example in *exps/customed_example/v++_SwinTiny_example.py*, please config the OVIS in the *get_data_loader* and *get_eval_loader* according to your own dataset. Remember to change the category information in the [evaluator](https://github.com/YuHengsss/YOLOV/blob/98ade28ce975291023be947b7d5d57b05f9600ba/yolox/evaluators/vid_evaluator_v2.py#L41). 3. Initialize the YOLOV or YOLOV++ with finetuned weights obtained by Step 1. You may adjust the hyperparameters such as [proposal numbers](https://github.com/YuHengsss/YOLOV/blob/8873e06cac9912c60c31ca2ef3061d0bfe5b2f36/exps/yolov_ovis/yolovs_ovis_75_75_750.py#L56) according to your dataset for getting better performance: ```shell python tools/vid_train.py -f exps/customed_example/v++_SwinTiny_example.py -c [path to your weights] --fp16 ``` Note that the batch size when training video detector is determined by the lframe and gframe, refer to this [line](https://github.com/YuHengsss/YOLOV/blob/98ade28ce975291023be947b7d5d57b05f9600ba/exps/yolov/yolov_base.py#L320). You can adjust the batch size according to your GPU memory. However, a very small batch size (<4) may lead to poor performance.

Acknowledgements

Expand

* [https://github.com/Megvii-BaseDetection/YOLOX](https://github.com/Megvii-BaseDetection/YOLOX) * [https://github.com/AlbertoSabater/Robust-and-efficient-post-processing-for-video-object-detection](https://github.com/AlbertoSabater/Robust-and-efficient-post-processing-for-video-object-detection)

Cite YOLOV and YOLOV++

If YOLOV series are helpful for your research, please cite the following paper:


@article{shi2024yolovpp,
      title={Practical Video Object Detection via Feature Selection and Aggregation}, 
      author={Yuheng Shi and Tong Zhang and Xiaojie Guo},
      journal={arXiv preprint arXiv:2407.19650},
      year={2024},
}

@article{shi2022yolov,
  title={YOLOV: Making Still Image Object Detectors Great at Video Object Detection},
  author={Shi, Yuheng and Wang, Naiyan and Guo, Xiaojie},
  journal={arXiv preprint arXiv:2208.09686},
  year={2022}
}

YuHengsss / YOLOV

readme