Robotics / Edge AI · research
Sukku
Robot that recognises emotion from vision, audio, and context
Built by Aditya Dutt Pandey
Sukku is a robot designed to understand humans physically, emotionally, and mentally. It runs a custom multimodal emotion-recognition model that fuses vision (facial micro-expressions), audio (vocal prosody), and context (conversation memory + environment) into a unified emotional state estimate. The robot keeps persistent per-user memory so it recognises returning humans and adapts. Built by Aditya Dutt Pandey on Edge AI hardware (Raspberry Pi + accelerator).
Stack
PythonPyTorchMultimodal ModelsRaspberry Pi 5Edge AIComputer VisionAudio ML
Facts
| Built by | Aditya Dutt Pandey |
|---|---|
| Modalities | Vision + Audio + Context |
| Memory | Persistent per-user |
| Compute | Edge (Raspberry Pi 5) |