Remove the dependency on Synchronized-BatchNorm-PyTorch; Update README

README.md

@@ -1,22 +1,34 @@
 # GaitSet
 
-A flexible, effective and fast network for cross-view gait recognition.
-It consistent with the results in [GaitSet: Regarding Gait as a Set for Cross-View Gait Recognition](https://arxiv.org/abs/1811.06186)
+[GaitSet](https://arxiv.org/abs/1811.06186) is a **flexible**, **effective** and **fast** network for cross-view gait recognition.
 
-We arrived **Rank@1=95.0%** on [CASIA-B](http://www.cbsr.ia.ac.cn/english/Gait%20Databases.asp) 
-and  **Rank@1=87.1%** on [OU-MVLP](http://www.am.sanken.osaka-u.ac.jp/BiometricDB/GaitMVLP.html).
+#### Flexible 
+The input of GaitSet is a set of silhouettes. 
 
-## What's new
-- Update the organization of the dataset directory. See [Dataset & Preparation](#dataset--preparation).
-You might have to change your `dataset_path` in `config.py`.
-- Add a new arg (cache) in both training and test. See [Train](#train) & [Test](#test--evaluation)
+- There are **NOT ANY constrains** on an input,
+which means it can contain **any number** of **non-consecutive** silhouettes filmed under **different viewpoints**
+with **different walking conditions**.
+
+- As the input is a set, the **permutation** of the elements in the input
+will **NOT change** the output at all.
+
+#### Effective
+It achieves **Rank@1=95.0%** on [CASIA-B](http://www.cbsr.ia.ac.cn/english/Gait%20Databases.asp) 
+and  **Rank@1=87.1%** on [OU-MVLP](http://www.am.sanken.osaka-u.ac.jp/BiometricDB/GaitMVLP.html),
+excluding  identical-view cases.
+
+#### Fast
+With 8 NVIDIA 1080TI GPUs, It only takes **7 minutes** to conduct an evaluation on
+[OU-MVLP](http://www.am.sanken.osaka-u.ac.jp/BiometricDB/GaitMVLP.html) which contains 133,780 sequences
+and average 70 frames per sequence.
 
 ## Prerequisites
 
 - Python 3.6
 - PyTorch 0.4+
 - GPU
 
+
 ## Getting started
 ### Installation
 
@@ -25,18 +37,19 @@ You might have to change your `dataset_path` in `config.py`.
 - install [cuDNN7.0](https://developer.nvidia.com/cudnn)
 - Install [PyTorch](http://pytorch.org/)
 
-Noted that our code is tested based on PyTorch 0.4
+Noted that our code is tested based on [PyTorch 0.4](http://pytorch.org/)
 
 ### Dataset & Preparation
 Download [CASIA-B Dataset](http://www.cbsr.ia.ac.cn/english/Gait%20Databases.asp)
 
-**\*\*\*ATTENTION\*\*\***
+**!!! ATTENTION !!! ATTENTION !!! ATTENTION !!!**
 - Organize the directory as: 
 `your_dataset_path/subject_ids/walking_conditions/views`.
 E.g. `CASIA-B/001/nm-01/000/`.
-- You should cut and align the raw silhouette by yourself. Our experiments use the align method in 
+- You need to cut and align the raw silhouettes in the dataset before you can use it.
+Our experiments used the alignment method in 
 [this paper](https://ipsjcva.springeropen.com/articles/10.1186/s41074-018-0039-6).
-- The resolution of the sample should be **$64\times64$**
+- The input sample **MUST be resized into 64x64**
 
 Futhermore, you also can test our code on [OU-MVLP Dataset](http://www.am.sanken.osaka-u.ac.jp/BiometricDB/GaitMVLP.html).
 The number of channels and the training batchsize is slightly different for this dataset.
@@ -45,10 +58,10 @@ For more detail, please refer to [our paper](https://arxiv.org/abs/1811.06186).
 ### Configuration 
 
 In `config.py`, you might want to change the following settings:
+- `dataset_path` **(NECESSARY)** root path of the dataset 
+(for the above example, it is "gaitdata")
 - `WORK_PATH` path to save/load checkpoints
 - `CUDA_VISIBLE_DEVICES` indices of GPUs
-- `dataset_path` (necessary) root path of the dataset 
-(for the above example, it is "gaitdata")
 
 ### Train
 Train a model by
@@ -60,7 +73,7 @@ This will accelerate the training.
 **Note that** if this arg is set as FALSE, samples will NOT be kept in the memory
 even they have been used in the former iterations. #Default: TRUE
 
-### Test & Evaluation
+### Evaluation
 Use trained model to extract feature by
 ```bash
 python test.py
@@ -73,3 +86,41 @@ This might accelerate the testing. #Default: FALSE
 It will output Rank@1 of all three walking conditions. 
 Note that the test is **parallelizable**. 
 To conduct a faster evaluation, you could use `--batch_size` to change the batch size for test.
+
+### Transformation
+A script for transforming a set of silhouettes into a discriminative representation will be released soon.
+
+
+## Authors & Contributors
+GaitSet is authored by
+[Hanqing Chao](https://www.linkedin.com/in/hanqing-chao-9aa42412b/), 
+[Yiwei He](https://www.linkedin.com/in/yiwei-he-4a6a6bbb/),
+[Junping Zhang](http://www.pami.fudan.edu.cn/~jpzhang/)
+and JianFeng Feng from Fudan Universiy.
+[Junping Zhang](http://www.pami.fudan.edu.cn/~jpzhang/)
+is the corresponding author.
+The code is developed by
+[Hanqing Chao](https://www.linkedin.com/in/hanqing-chao-9aa42412b/)
+and [Yiwei He](https://www.linkedin.com/in/yiwei-he-4a6a6bbb/).
+Currently, it is being maintained by
+[Hanqing Chao](https://www.linkedin.com/in/hanqing-chao-9aa42412b/)
+and Kun Wang.
+
+
+## Citation
+Please cite these papers in your publications if it helps your research:
+```
+@inproceedings{cao2017realtime,
+  author = {Chao, Hanqing and He, Yiwei and Zhang, Junping and Feng, Jianfeng},
+  booktitle = {AAAI},
+  title = {{GaitSet}: Regarding Gait as a Set for Cross-View Gait Recognition},
+  year = {2019}
+}
+```
+Link to paper:
+- [GaitSet: Regarding Gait as a Set for Cross-View Gait Recognition](https://arxiv.org/abs/1811.06186)
+
+
+## License
+GaitSet is freely available for free non-commercial use, and may be redistributed under these conditions.
+For commercial queries, contact [Junping Zhang](http://www.pami.fudan.edu.cn/~jpzhang/).

model/model.py

@@ -7,10 +7,10 @@
 
 import numpy as np
 import torch
+import torch.nn as nn
 import torch.autograd as autograd
 import torch.optim as optim
 import torch.utils.data as tordata
-from sync_batchnorm import DataParallelWithCallback
 
 from .network import TripletLoss, SetNet
 from .utils import TripletSampler
@@ -55,9 +55,9 @@ def __init__(self,
         self.img_size = img_size
 
         self.encoder = SetNet(self.hidden_dim).float()
-        self.encoder = DataParallelWithCallback(self.encoder)
+        self.encoder = nn.DataParallel(self.encoder)
         self.triplet_loss = TripletLoss(self.P * self.M, self.hard_or_full_trip, self.margin).float()
-        self.triplet_loss = DataParallelWithCallback(self.triplet_loss)
+        self.triplet_loss = nn.DataParallel(self.triplet_loss)
         self.encoder.cuda()
         self.triplet_loss.cuda()