11 accepted papers
Multimodal large language models (MLLMs) demonstrate strong video understanding by attending to visual tokens relevant to instructions. To exploit this for training-free localization, we cast video reasoning segmentation as video QA and extract attention maps via rollout. Since raw maps are too nois…
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 insigh…
In a retrieval system, simultaneously achieving search accuracy and efficiency is inherently challenging. This challenge is particularly pronounced in partially relevant video retrieval (PRVR), where incorporating more diverse context representations at varying temporal scales for each video enhance…
"Video action detection (VAD) aims to detect actors and classify their actions in a video. We figure that VAD suffers more from classification rather than localization of actors. Hence, we analyze how prevailing methods form features for classification and find that they prioritize actor regions, ye…
Most of the existing Video Anomaly Detection (VAD) studies have been conducted within single-domain learning, where training and evaluation are performed on a single dataset. However, the criteria for abnormal events differ across VAD datasets, making it problematic to apply a single-domain model to…
Temporal action detection aims to predict the time intervals and the classes of action instances in the video. Despite the promising performance, existing two-stream models exhibit slow inference speed due to their reliance on computationally expensive optical flow. In this paper, we introduce a dec…
Data augmentation has recently emerged as an essential component of modern training recipes for visual recognition tasks. However, data augmentation for video recognition has been rarely explored despite its effectiveness. Few existing augmentation recipes for video recognition naively extend the im…
Recent self-supervised video representation learning methods focus on maximizing the similarity between multiple augmented views from the same video and largely rely on the quality of generated views. However, most existing methods lack a mechanism to prevent representation learning from bias toward…
Most existing multi-person tracking approaches rely on appearance based re-identification (re-ID) to resolve fragmented tracklets. However, simply using appearance information could be insufficient for videos containing severe pose changes, such as sports or dance videos. With the goal of learning p…
Person re-identification (re-ID) is the problem of visually identifying a person given a database of identities. In this work, we focus on image-to-video re-ID which compares a single query image to videos in the gallery. The main challenge is the asymmetry association of an image and a video, and o…
A major obstacle in teaching machines to understand videos is the lack of training data, as creating temporal annotations for long videos requires a huge amount of human effort. To this end, we introduce a new large-scale baseball video dataset called the BBDB, which is produced semi-automatically b…