EventCoT: Event-centric Video Chain-of-thought for Reasoning Temporal Localization
2026-07-06 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial Intelligence
AI summaryⓘ
The authors introduce EventCoT, a new method to answer questions about videos by not only providing the answer but also the exact time segment in the video where the answer is found. Their approach breaks the video into meaningful events to focus on relevant parts, making it more efficient. EventCoT matches its reasoning with the visual content to accurately locate the time intervals. The authors show that their method outperforms previous ones on a key dataset and works well on another related video question answering test without extra training.
Temporal LocalizationVideo Question AnsweringEvent TokenizationChain of Thought ReasoningVisual EmbeddingsActivityNet-RTLZero-shot LearningGrounded Reasoning
Authors
Youngkil Song, Yoonjae Baek, Dongwon Kim, Inho Kim, Dongkeun Kim, Suha Kwak
Abstract
Reasoning temporal localization (RTL) requires a model to generate an answer that itself contains the time interval supporting it, so high-level reasoning and precise temporal grounding must be produced jointly in a single response. To tackle this challenging task, we propose the first event-centric video chain-of-thought framework, dubbed EventCoT. EventCoT first performs event-centric tokenization of the input video to convert it into compact event tokens, enabling efficient identification of question-relevant events. It then reasons within the identified events to generate the answer, grounding the time interval via embedding matching that aligns placeholder tokens with visual embeddings. EventCoT achieves state-of-the-art results on ActivityNet-RTL for reasoning temporal localization while using substantially fewer visual tokens than previous work. To verify its general performance, we further evaluate EventCoT on the grounded video question answering benchmark ReXTime, where it attains strong zero-shot results.