Real-Time ISL Translation Model -- 2
Budget: ₹1,500 – ₹5,000 INR
I need highly accurate deep-learning model that translates Indian Sign Language gestures into words or phrases. The core outcome I’m after is translation accuracy; latency can be modest as long as it feels near real time for a webcam feed.
Key points
• Input: live video through camera feed or MediaPipe/OpenPose keypoints.
• Output: English word/phrase predictions rendered on a simple web page.
• Platform: Web deployment only—TensorFlow JS or another browser-friendly approach is ideal.
• Scope: deliver a working proof-of-concept with clear code, minimal UI, and instructions so I can test the model end to end.
• Data: you may leverage public ISL datasets
• Deadline: ASAP—please outline a realistic schedule for an initial demo and final hand-off.
If you have prior sign-language projects, especially with keypoint-based pipelines or sign language translation models, mention them—I’ll prioritize proven experience.
Key points
• Input: live video through camera feed or MediaPipe/OpenPose keypoints.
• Output: English word/phrase predictions rendered on a simple web page.
• Platform: Web deployment only—TensorFlow JS or another browser-friendly approach is ideal.
• Scope: deliver a working proof-of-concept with clear code, minimal UI, and instructions so I can test the model end to end.
• Data: you may leverage public ISL datasets
• Deadline: ASAP—please outline a realistic schedule for an initial demo and final hand-off.
If you have prior sign-language projects, especially with keypoint-based pipelines or sign language translation models, mention them—I’ll prioritize proven experience.