2025 IEEE LOW-POWER COMPUTER VISION CHALLENGE

Track 2 - Open-vocabulary segmentation with text prompt