Request: AI-based Deaf Sign Language Project using RealSense
-
Hi Yasodeafs The RealSense D435 camera model is likely to meet your requirements, as long as the person is standing 1.5 meters or more from the camera. Its field of vision is wide enough to capture a person standing in front of it and hand-gesturing.
The RealSense SDK software does not have face and gesture detection built in, and if your project is non-commercial then a RealSense-compatible commercial gesture and face tracking solution may not meet your needs.
It is possible to train a recognition dataset yourself using RealSense cameras, as described by an Intel seminar video on YouTube, and therefore comply with your non-commercial goals.
https://www.youtube.com/watch?v=VSHDyUXSNqY
In regard to RealSense and AI accessibility, the case-study article about Lumen may be a helpful reference.
https://www.intelrealsense.com/3d-vision-for-the-visually-impaired/
-
Hi MartyG and RealSense Team,
Thank you so much for your detailed and respectful response.
I truly appreciate your suggestions regarding the D435 model and the possibility of training a recognition dataset for non-commercial purposes. This aligns well with the mission of our ChatDEAF project, which is focused on AI-driven accessibility for the Deaf community.
We will definitely review the resources you shared, including the Lumen case study and the YouTube seminar. If we have further technical questions during the training phase, we would be honored to reach out again.
Thanks again for supporting accessibility and innovation.
Kind regards,
Yasin Şimşek
Founder – ChatDEAF -
Dear Marty,
Thank you for your positive response and support.
I’m currently advancing ChatDEAF’s prototype stage and your openness to future technical questions means a lot.
Once we finalize the team and move deeper into training, I hope to share updates with the RealSense community.Warm regards,
Yasin Şimşek
Founder – ChatDEAF
Please sign in to leave a comment.
Comments
4 comments