Vision Language Model: Scene Description on SL2610-SeriesDescribe what the camera sees with the LFM2-VL vision-language model on the SL2610 NPU.