| Pedestrian Attribute Analysis | StrongBaseline | ma: 94.86 | Per Person 2ms | [Download Link](https://bj.bcebos.com/v1/paddledet/models/pipeline/strongbaseline_r50_30e_pa100k.tar) |
| High-Precision Model | PP-HGNet_small | mA: 95.4 | per person 1.54ms | [Download](https://bj.bcebos.com/v1/paddledet/models/pipeline/PPLCNet_x1_0_person_attribute_945_infer.tar) |
| Fast Model | PP-LCNet_x1_0 | mA: 94.5 | per person 0.54ms | [Download](https://bj.bcebos.com/v1/paddledet/models/pipeline/PPLCNet_x1_0_person_attribute_945_infer.tar) |
| Balanced Model | PP-HGNet_tiny | mA: 95.2 | per person 1.14ms | [Download](https://bj.bcebos.com/v1/paddledet/models/pipeline/PPHGNet_tiny_person_attribute_952_infer.tar) |
1. The precision of detection/ tracking models is obtained by training and testing on the dataset consist of [MOT17](https://motchallenge.net/)，[CrowdHuman](http://www.crowdhuman.org/)，[HIEVE](http://humaninevents.org/) and some business data.
2. The precision of pedestiran attribute analysis is obtained by training and testing on the dataset consist of [PA100k](https://github.com/xh-liu/HydraPlus-Net#pa-100k-dataset)，[RAPv2](http://www.rapdataset.com/rapv2.html)，[PETA](http://mmlab.ie.cuhk.edu.hk/projects/PETA.html) and some business data.
3. The inference speed is T4, the speed of using TensorRT FP16.
1. The precision of pedestiran attribute analysis is obtained by training and testing on the dataset consist of [PA100k](https://github.com/xh-liu/HydraPlus-Net#pa-100k-dataset)，[RAPv2](http://www.rapdataset.com/rapv2.html)，[PETA](http://mmlab.ie.cuhk.edu.hk/projects/PETA.html) and some business data.
2. The inference speed is V100, the speed of using TensorRT FP16.
## Instruction
...
...
@@ -70,7 +70,7 @@ Data Source and Copyright：Skyinfor Technology. Thanks for the provision of act
- Boots: Yes; No
```
4. The model adopted in the attribute recognition is [StrongBaseline](https://arxiv.org/pdf/2107.03576.pdf), where the structure is the multi-class network structure based on ResNet50, and Weighted BCE loss and EMA are introduced for effect optimization.
4. The model adopted in the attribute recognition is [StrongBaseline](https://arxiv.org/pdf/2107.03576.pdf), where the structure is the multi-class network structure based on PP-HGNet、PP-LCNet, and Weighted BCE loss is introduced for effect optimization.
@@ -49,7 +49,7 @@ capture the target in the original image according to bbox——│
make the IDs cluster together and rearrange them
```
2. The model solution is [reid-centroids](https://github.com/mikwieczorek/centroids-reid), with ResNet50 as the backbone. It is worth noting that the solution employs different features of the same ID to enhance the similarity.
2. The model solution is [reid-strong-baseline](https://github.com/michuanhaohao/reid-strong-baseline), with ResNet50 as the backbone.
Under the above circumstances, the REID model used in MTMCT integrates open-source datasets and compresses model features to 128-dimensional features to optimize the generalization. In this way, the actual generalization result becomes much better.
...
...
@@ -74,11 +74,21 @@ Under the above circumstances, the REID model used in MTMCT integrates open-sour
## Reference
```
@article{Wieczorek2021OnTU,
title={On the Unreasonable Effectiveness of Centroids in Image Retrieval},
author={Mikolaj Wieczorek and Barbara Rychalska and Jacek Dabrowski},
journal={ArXiv},
year={2021},
volume={abs/2104.13643}
@InProceedings{Luo_2019_CVPR_Workshops,
author = {Luo, Hao and Gu, Youzhi and Liao, Xingyu and Lai, Shenqi and Jiang, Wei},
title = {Bag of Tricks and a Strong Baseline for Deep Person Re-Identification},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2019}
}
@ARTICLE{Luo_2019_Strong_TMM,
author={H. {Luo} and W. {Jiang} and Y. {Gu} and F. {Liu} and X. {Liao} and S. {Lai} and J. {Gu}},
journal={IEEE Transactions on Multimedia},
title={A Strong Baseline and Batch Normalization Neck for Deep Person Re-identification},
