Overview

This project note records our result in the 2026 IEEE Low Power Computer Vision Challenge (LPCVC), held as part of the Efficient Deep Learning for Computer Vision workshop at CVPR 2026.

In the Action Recognition in Video track, our team TEAM-ALPHA placed 2nd. The submitted model was evaluated on the Qualcomm Dragonwing IQ-9075 EVK through Qualcomm AI Hub.

Qualcomm Dragonwing chip
  • Model: DecomposedVideoMAE (ViT-B, 112px, float)
  • Latency: 19.322 ms on Dragonwing IQ-9075 EVK
  • LB Accuracy: 96.776%

Reference

External Link LPCVC 2026 Winners lpcv.ai lpcv lpcv.ai lpcv.ai
External Link ECV @ CVPR 2026 Workshop ecv-workshop.github.io ecv-workshop ecv-workshop.github.io ecv-workshop.github.io

Solutions (public code)

The implementation has been published here:

LPCVC_2026_Track2_team-alpha Public Code LPCVC_2026_Track2_team-alphatasuku-takagi