Mexican Sign Language's Dactylology and Ten First Numbers - Labeled images and videos. From person #1 to #5
Description
The dataset comprises edited recordings of Mexican sign language's Dactylology (29 signs) and Ten First Numbers (from 1 to 10), including static and continuous signs accordingly from person 1 to person 5. The edited recordings are organized for easy access and management. Edited videos and screenshots of static signs are labeled in their file with corresponding sign language representations and stored in a consistent order per person, the number of the cycle of recording, and per hand. Static sign images can be exported in PASCAL VOC format with XML annotations too. The dataset is designed to facilitate feature extraction and further analysis in Mexican sign language recognition research.
Files
Steps to reproduce
1. Obtain the raw recordings by capturing the recordings of sign language gestures, both static and continuous signs, ensuring an accurate representation of each sign per cycle of dactylology and the first ten numbers. 2. Organize the recordings by storing the raw recordings in a suitable directory structure for easy access and management during the editing process. (Example: Google Drive). 3. Prepare the editing software by installing video editing software capable of handling both continuous and static sign recordings. Ensure the software supports precise control over the start and end times of continuous signs and allows capturing screenshots of static signs. Alternatively, you can use a smartphone to perform this task. 4. Edit continuous sign videos and adjust the start and end times to accurately represent each gesture. This step ensures that each continuous sign has the appropriate duration and captures the complete gesture cycle. Save each edited video as a video file with the same format (.mp4, or .mov), and the corresponding continuous sign language representation in the name of the file. 5. For each static sign, navigate to the corresponding frame in the video and capture a screenshot. Save each screenshot as an image file with the same format (.jpg, or .jpeg) and the corresponding static sign language representation in the name of the file. This labeling in the name of the file helps identify each sign easily and enables further analysis. 6. [Optional] Export static sign images in PASCAL VOC format: Use a tool like LabelImg to tag the static sign images in the PASCAL VOC format. Open each image in LabelImg, draw bounding boxes around the static signs, and assign the corresponding label. Save the annotations as XML files, ensuring they adhere to the PASCAL VOC format standards. 7. Store the edited videos, labeled images, and XML files in a consistent order per person, then per cycle of recording, and finally per hand. The order of the folders and the way it’s organized will help in the way of extracting the features in a repository (Example: Google Drive) in future stages.