crosstyan/OpenGait
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: add pose index files using guide in datasets/OUMVLP/README.md

Browse Source
This commit is contained in:
Ahzyuan
2025-09-06 21:18:31 +08:00
parent cedced825a
commit 62b8915b26
1 changed files with 29 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
datasets/OUMVLP/README.md
+29 -4
View File
@@ -37,7 +37,7 @@ python datasets/OUMVLP/extractor.py --input_path Path_of_OUMVLP-base --output_pa
```
Step3-1 : To rearrange directory of OUMVLP dataset(for silhouette), turning to id-type-view structure, Run
```
python datasets/OUMVLP/rearrange_OUMVLP.py --input_path Path_of_OUMVLP-raw --output_path Path_of_OUMVLP-rearranged
python datasets/OUMVLP/rearrange_OUMVLP.py --input_path Path_of_OUMVLP-raw --output_path Path_of_OUMVLP-silu-rearranged
```
Step3-2 : To rearrange directory of OUMVLP dataset(for pose), turning to id-type-view structure, Run
```
@@ -46,12 +46,37 @@ python datasets/OUMVLP/rearrange_OUMVLP_pose.py --input_path Path_of_OUMVLP-pose
Step4-1: Transforming images to pickle file, run
```
python datasets/pretreatment.py --input_path Path_of_OUMVLP-rearranged --output_path Path_of_OUMVLP-pkl
python datasets/pretreatment.py --input_path Path_of_OUMVLP-silu-rearranged --output_path Path_of_OUMVLP-pkl
```
Step4-2: Transforming pose txts to pickle file, run
> [!IMPORTANT]
> Before extracting pose pkls, **you need to possess the pose selection index files** ([Why](https://github.com/ShiqiYu/OpenGait/pull/280)). Here are two ways to get it:
> 1. `Approach 1`: Directly download it:
> - Open [Download Link](https://drive.google.com/drive/folders/1gkXdrVtNuGbU5wd8lWoPfAo_qYpokm52?usp=sharing), choose `AlphaPose` or `OpenPose` version
> - Find a suitable location to unzip it, like `<somewhere>/OUMVLP/Pose/match_idx`.
> - Move the zip file into the `match_idx` dir and unzip it there.
> - You will finally get the index root: `<somewhere>/OUMVLP/Pose/match_idx/AlphaPose`
> *(Here we take `AlphaPose` version as an example, this path is what we call `Path_of_OUMVLP-pose-index` below)*
>
> 2. `Approach 2`: Run the following command to generate it by yourself (**rearranged silhouette dataset is needed**):
>
> ```bash
> python datasets/OUMVLP/pose_index_extractor.py \
> -p Path_of_OUMVLP-pose-rearranged \
> -s Path_of_OUMVLP-silu-rearranged \
> -o Path_of_OUMVLP-pose-index
> ```
```bash
python datasets/pretreatment.py \
--input_path Path_of_OUMVLP-pose-rearranged \
--output_path Path_of_OUMVLP-pose-pkl \
--pose \
--dataset OUMVLP \
--oumvlp_index_dir Path_of_OUMVLP-pose-index
```
python datasets/pretreatment.py --input_path Path_of_OUMVLP-pose-rearranged --output_path Path_of_OUMVLP-pose-pkl --pose --dataset OUMVLP --oumvlp_rearrange_silu_path Path_of_OUMVLP-rearranged-silu-dataset
```
gernerate the 17 Number of Pose Points Format from 18 Number of Pose Points
```
python datasets/OUMVLP/rearrange_OUMVLP_pose.py --input_path Path_of_OUMVLP-pose18 --output_path Path_of_OUMVLP-pose17