Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
The paper introduces MarkIt, a training-free framework that enhances video temporal grounding by converting input videos into query-conditioned marked videos for Vid-LLMs. MarkIt leverages a query-to-mask grounding bridge (Q2M-Bridge) that derives subject tags from queries and maps them to instance masks using open-vocabulary segmentation, embedding semantic instance markers and frame indices into each frame. This approach provides explicit visual cues for temporal reasoning, leading to state-of-the-art results on moment retrieval and highlight detection benchmarks without requiring Vid-LLM fine-tuning.
Forget fine-tuning: MarkIt's training-free visual markers give Vid-LLMs superhuman temporal grounding abilities.
Video temporal grounding (VTG) aims to localize the start and end timestamps of the event described by a given query within an untrimmed video. Despite the strong open-world video understanding and recognition ability of video language large models (Vid-LLMs), outputting precise temporal grounding information remains challenging, since explicit temporal cues are scarce in untrimmed videos, and query-relevant entities are hard to track consistently across the video timeline. In this paper, we present \MarkIt{}, a training-free framework that transforms an input video into a query-conditioned marked video, which empowers Vid-LLMs to generate more reliable temporal localization predictions. The core component of \MarkIt{} is an annotation-free query-to-mask grounding bridge (Q2M-Bridge). Given a natural-language query, it automatically derives a compact set of canonical subject tags through linguistic parsing and normalization, then maps these tags to query-conditioned instance masks using text-conditioned open-vocabulary segmentation. The bridge also embeds lightweight semantic instance markers and a persistent frame index into each frame, effectively transforming long-range temporal reasoning into explicit visual cues for Vid-LLMs. \MarkIt{} adopts an inference-time plug-and-play design, needs no modifications to Vid-LLM weights, and is fully compatible with supervised fine-tuning. Experiments conducted on multiple mainstream moment retrieval and highlight detection benchmarks demonstrate that \MarkIt {} achieves state-of-the-art results, delivering consistent temporal grounding improvements across a wide range of existing models.