"Tsun2x team on AMD AI League Hackathon"
Team Idea
Combining low-latency GStreamer RTSP ingestion with synthetic trajectory simulation creates a compute-lean benchmarking framework for TAVIDM. The target is to minimize hardware overhead by running heavy neural network detections only when necessary while keeping tracking smooth using low-cost state estimators (Kalman Filters / Optical Flow).
Team Members
.png&w=256&q=75)
Unknown Hacker
markosumo1597794
Submission
Submission draft in progress