yasodeafs/chatdeaf-isl-gestures

ChatDEAF ISL Dataset-Initial README and metadata update

---

license: cc-by-4.0 pretty_name: ChatDEAF ISL Gesture Dataset tags:

• sign-language

• accessibility

• isl

• gesture-recognition

• multimodal

• chatdeaf

task_categories:

• image-classification

size_categories:

• 10<n<100

language: - en

tags:

• sign-language
• isl
• international-sign
• gestures
• accessibility
• chatdeaf
• image-classification
• visual-language

{ "belt": "Gesture for 'belt' in ISL.", "glasses": "Gesture for 'glasses' in ISL.", "tree": "Gesture for 'tree' in ISL.", "wow": "Gesture expressing surprise or excitement.", "bravo": "Clapping gesture for approval." }

ChatDEAF ISL Gesture Dataset

The ChatDEAF ISL Gesture Dataset is a reference collection of International Sign Language (ISL) gestures. It includes images, gesture labels, and written descriptions designed to support AI models in accessibility, sign recognition, and gesture-text classification.

Contents

• 18 sample ISL gestures
• English word labels
• Gesture descriptions
• Image file references

Purpose

This dataset supports accessibility research and aims to enhance multimodal AI systems (like GPT-4) in understanding visual sign languages.

License

This dataset is released under the CC BY 4.0 license.

ChatDEAF: giving voice to the silence.

ChatDEAF ISL Gesture Dataset

This is the first public dataset of International Sign Language (ISL) gestures created as part of the ChatDEAF Project. The dataset is intended to support multimodal AI models and improve accessibility in communication for the global deaf community.

Dataset Details

• Total Samples: 18 labeled ISL gesture entries
• Fields:
  • Word: English meaning of the gesture
  • ISL Sign Description: Text description of the sign gesture
  • Image: Corresponding photo or image file (to be expanded in future versions)

License

CC-BY-4.0 — This dataset is open for academic, research, and responsible AI usage with proper attribution.

Purpose

This dataset supports gesture recognition training, sign language understanding models, and accessibility technology research. It's a foundational step in training AI to understand international sign language.

Author

• AI gesture recognition training
• Accessibility research
• Multimodal AI systems (text + image)
• Sign language learning tools

About ChatDEAF

The ChatDEAF Project is a grassroots initiative led by Yasin Şimşek, a deaf researcher from Türkiye. This dataset marks the beginning of a larger mission to build open, diverse, and inclusive sign language AI tools.

Future Plans

• Expand to 100+ ISL gestures
• Add Turkish Sign Language (TİD) support
• Create a HuggingFace Space demo for live ISL gesture recognition

---

ChatDEAF ISL Gesture Dataset

This dataset contains labeled images and descriptions of International Sign Language (ISL) gestures. All files are located under the /data directory.

Folder Structure

chatdeaf-isl-gestures/ ├── README.md ├── data/ │ ├── ChatDEAF_ISL_Dataset_Proposal.csv │ ├── police-1.jpg │ ├── pray-1.jpg │ └── ...

Fields

• Word: The English translation of the sign
• ISL Sign Description: Textual description of the gesture
• Image: Visual photo of the sign (JPEG)

License: CC BY 4.0

Purpose

To support gesture recognition, sign language translation models, and accessibility research.

ChatDEAF is an open dataset of International Sign Language (ISL) gestures created by Yasin Şimşek, a Deaf sign language user from Türkiye.

The dataset includes labeled gesture names, visual references, and gesture descriptions in English. It aims to support AI research in accessibility, sign recognition, and multimodal understanding.

This is the first version (v1) with 18 sample gestures. The dataset will be expanded in future releases.

[email protected] thank you.