Multimodal AI

Vision-Language Models, Multimodal architectures, and Video Understanding

1.

Image Captioning WIP

Architectures for generating natural language descriptions of images

2.

Visual Question Answering (VQA) WIP

Answering natural language questions about visual content

3.

Video Understanding WIP

Processing, analyzing, and extracting spatio-temporal features from video data

4.

Unified Latent Architectures WIP

Cross-domain competence through value-aligned latent representations without full model-based planning