The combination of strong visual backbones and Large Language Model (LLM) reasoning
has led to Large Multimodal Models (LMMs) becoming the current standard for a wide range …

Vision and Language (VL) models have demonstrated remarkable zero-shot performance in
a variety of tasks. However, some aspects of complex language understanding still remain a …

Vision and Language (VL) models offer an effective method for aligning representation
spaces of images and text allowing for numerous applications such as cross-modal retrieval …

We introduce an object-aware decoder for improving the performance of spatio-temporal
representations on ego-centric videos. The key idea is to enhance object-awareness during …

Vision and language models (VLMs) have demonstrated remarkable zero-shot (ZS)
performance in a variety of tasks. However, recent works have shown that even the best …

We introduce a new attention mechanism dubbed structural self-attention (StructSA) that
leverages rich correlation patterns naturally emerging in key-query interactions of attention …

Action recognition models have achieved impressive results by incorporating scene-level
annotations, such as objects, their relations, 3D structure, and more. However, obtaining …

The recent success of interleaved Large Multimodal Models (LMMs) in few-shot learning
suggests that in-context learning (ICL) with many examples can be promising for learning …

Different video understanding tasks are typically treated in isolation, and even with distinct
types of curated data (eg, classifying sports in one dataset, tracking animals in another) …

This paper presents Action-SGFA, a novel action feature alignment approach to learn unified
joint embeddings across four action modalities incorporating scene graph (SG) …