PFN unveils PLaMo-VL visual-language models for drones and robots

Preferred Networks released PLaMo-VL, visual-language models for autonomous devices, in 8B and 2B sizes.

The models are designed to run on edge devices and improve VQA and Visual Grounding capabilities.

PFN applied the models to factory task classification and power-plant anomaly detection and reported benchmark gains.

PFN received the GENIAC model award and plans further field tests and additional training with real-world data.