Dataset Viewer
The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
step 1. Origin Data Download
Keypoints
pre-extracted by HRNet. Please download them and place them under OriginData/(Phoenix-2014T or CSL-Daily).
step 2. Pose 133 to 75 (Adjusted keypoints from 133 to 75 and eliminated some noisy frames. )
python CSL-Daily-Clean.py
python Phoenix-2014T-Clean.py
step 3. 2D to 3D Pose (Use the model to generate keypoint depth information and normalize the data. This step takes a long time.)
python pipeline.py --input "133to75/CSL-Daily/CSL-Daily.dev" --output "SLRTData/CSL-Daily/CSL-Daily.dev"
python pipeline.py --input "133to75/CSL-Daily/CSL-Daily.test" --output "SLRTData/CSL-Daily/CSL-Daily.test"
python pipeline.py --input "133to75/CSL-Daily/CSL-Daily.train" --output "SLRTData/CSL-Daily/CSL-Daily.train"
python pipeline.py --input "133to75/Phoenix-2014T/Phoenix-2014T.dev" --output "SLRTData/Phoenix-2014T/Phoenix-2014T.dev"
python pipeline.py --input "133to75/Phoenix-2014T/Phoenix-2014T.test" --output "SLRTData/Phoenix-2014T/Phoenix-2014T.test"
python pipeline.py --input "133to75/Phoenix-2014T/Phoenix-2014T.train" --output "SLRTData/Phoenix-2014T/Phoenix-2014T.train"
step 4. Sign language production data processing
python CSL-Daily-SLP.py
python Phoenix-2014T-SLP.py
- Downloads last month
- 126