SpeedTuner/arin460-blip-visual-question-answering

BLIP Multimodal Visual Question Answering

ARIN 460 Unit 8 capstone deployment using Salesforce/blip-vqa-base.

The application accepts two modalities:

1. 1.An uploaded image
2. 2.A natural-language question

It returns the model's answer, confidence information, inference latency, and alternative predictions.

Responsible use

This demonstration is educational and should not be used for safety-critical decisions.