As the scale of data and models for video understanding rapidly expand, handling long-form video input in transformer-based models presents a practical challenge. Rather than resorting to input sampling or token dropping, which may result in information loss, token merging shows promising results when used in collaboration with transformers. However, the application of token merging for long-form video processing is not trivial. We begin with the premise that token merging should not rely solely on the similarity of video tokens; the saliency of tokens should also be considered. To address this, we explore various video token merging strategies for long-form video classification, starting with a simple extension of image token merging, moving to region-concentrated merging, and finally proposing a learnable video token merging (VTM) algorithm that dynamically merges tokens based on their saliency. Extensive experimental results show that we achieve better or comparable performances on the LVU, COIN, and Breakfast datasets. Moreover, our approach significantly reduces memory costs by 84% and boosts throughput by approximately 6.89 times compared to baseline algorithms.

Video-Language Models (VLMs), powered by the advancements in Large Language Models (LLMs), are charting new frontiers in video understanding. A pivotal challenge is the development of an effective method to encapsulate video content into a set of representative tokens to align with LLMs. In this work, we introduce Slot-VLM, a new framework designed to generate semantically decomposed video tokens, in terms of object-wise and event-wise visual representations, to facilitate LLM inference.

Learning a generalist embodied agent capable of completing multiple tasks poses challenges, primarily stemming from the scarcity of action-labeled robotic datasets. In contrast, a vast amount of human videos exist, capturing intricate tasks and interactions with the physical world. Promising prospects arise for utilizing actionless human videos for pre-training and transferring the knowledge to facilitate robot policy learning through limited robot demonstrations. However, it remains a challenge due to the domain gap between humans and robots. Moreover, it is difficult to extract useful information representing the dynamic world from human videos, because of its noisy and multimodal data structure.

Recent long-form video-language understanding benchmarks have driven progress in video large multimodal models (Video-LMMs). However, the scarcity of well-annotated long videos has left the training of hour-long Video-LMMs underexplored. To close this gap, we present VideoMarathon, a large-scale hour-long video instruction-following dataset. This dataset includes around 9,700 hours of long videos sourced from diverse domains, ranging from 3 to 60 minutes per video. Specifically, it contains 3.3M high-quality QA pairs, spanning six fundamental topics: temporality, spatiality, object, action, scene, and event. Compared to existing video instruction datasets, VideoMarathon significantly extends training video durations up to 1 hour, and supports 22 diverse tasks requiring both short- and long-term video comprehension. Building on VideoMarathon, we propose Hour-LLaVA, a powerful and efficient Video-LMM for hour-scale video-language modeling. It enables hour-long video training and inference at 1-FPS sampling by leveraging a memory augmentation module, which adaptively integrates question-relevant and spatiotemporally informative semantics from the cached full video context. In our experiments, Hour-LLaVA achieves the best performance on multiple representative long video-language benchmarks, demonstrating the high quality of the VideoMarathon dataset and the superiority of the Hour-LLaVA model.

Video LLMs suffer from temporal inconsistency: small shifts in frame timing can flip attention and suppress relevant frames. We trace this instability to the common extension of Rotary Position Embeddings to video through multimodal RoPE. The induced inverse Fourier time kernel exhibits frame-scale ripples that multiply adjacent frames by different factors, which perturbs attention that should otherwise be governed by the raw query key inner product. We present Phase Aggregated Smoothing (PAS), a simple, training-free mechanism that applies small opposed phase offsets across heads and then aggregates their outputs. PAS preserves the per-head spectrum magnitude, while the aggregation effectively smooths the temporal kernel and reduces phase sensitivity without changing the positional encoding structure. Our analysis shows that the RoPE rotated logit can be approximated as a content dot product scaled by a time kernel; smoothing this kernel yields Lipschitz stability of attention to small temporal shifts; multi phase averaging attenuates high frequency ripples while preserving per-head spectra under Nyquist-valid sampling. Experiments on multiple video understanding benchmarks under matched token budgets show consistent improvements with negligible computational overhead. PAS provides a plug and play upgrade for robust temporal encoding in Video LLMs.

Work done during an internship at Amazon Prime Video. Work done while at Amazon Prime Video.

Video-Language Models (VLMs), powered by the advancements in Large Language Models (LLMs), are charting new frontiers in video understanding.

Video large language models (LLMs) achieve strong video understanding by leveraging a large number of spatio-temporal tokens, but suffer from quadratic computational scaling with token count. To address this, we propose a training-free spatio-temporal token merging method, named STTM. Our key insight is to exploit local spatial and temporal redundancy in video data which has been overlooked in prior work. STTM first transforms each frame into multi-granular spatial tokens using a coarse-to-fine search over a quadtree structure, then performs directed pairwise merging across the temporal dimension. This decomposed merging approach outperforms existing token reduction methods across six video QA benchmarks. Notably, STTM achieves a 2$\times$ speed-up with only a 0.5% accuracy drop under a 50% token budget, and a 3$\times$ speed-up with just a 2% drop under a 30% budget. Moreover, STTM is query-agnostic, allowing KV cache reuse across different questions for the same video. The project page is available at https://www.jshyun.me/projects/sttm.

Humanoid robots, with their human-like form, are uniquely suited for interacting in environments built for people. However, enabling humanoids to reason, plan, and act in complex open-world settings remains a challenge. World models, models that predict the future outcome of a given action, can support these capabilities by serving as a dynamics model in long-horizon planning and generating synthetic data for policy learning. We introduce Humanoid World Models (HWM), a family of lightweight, open-source models that forecast future egocentric video conditioned on humanoid control tokens. We train two types of generative models, Masked Transformers and Flow-Matching, on 100 hours of humanoid demonstrations. Additionally, we explore architectural variants with different attention mechanisms and parameter-sharing strategies. Our parameter-sharing techniques reduce model size by 33-53% with minimal impact on performance or visual fidelity. HWMs are designed to be trained and deployed in practical academic and small-lab settings, such as 1-2 GPUs.