ST-VLM: Kinematic Instruction Tuning for Spatio-Temporal Reasoning in Vision-Language Models

• URL: http://arxiv.org/abs/2503.19355v2
• Date: Wed, 26 Mar 2025 05:32:54 GMT
• Title: ST-VLM: Kinematic Instruction Tuning for Spatio-Temporal Reasoning in Vision-Language Models
• Authors: Dohwan Ko, Sihyeon Kim, Yumin Suh, Vijay Kumar B. G, Minseo Yoon, Manmohan Chandraker, Hyunwoo J. Kim,
• Abstract summary: Vision-Language Models (Ms) struggle to analyze elements like traveled distance and speed of moving objects.<n>We construct a benchmark dataset referred to as STKit and ST-Bench.<n>We show that ST-VLM generalizes robustly across diverse domains and tasks.
• Score: 63.12671761097701
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Spatio-temporal reasoning is essential in understanding real-world environments in various fields, eg, autonomous driving and sports analytics. Recent advances have improved the spatial reasoning ability of Vision-Language Models (VLMs) by introducing large-scale data, but these models still struggle to analyze kinematic elements like traveled distance and speed of moving objects. To bridge this gap, we construct a spatio-temporal reasoning dataset and benchmark involving kinematic instruction tuning, referred to as STKit and STKit-Bench. They consist of real-world videos with 3D annotations, detailing object motion dynamics: traveled distance, speed, movement direction, inter-object distance comparisons, and relative movement direction. To further scale such data construction to videos without 3D labels, we propose an automatic pipeline to generate pseudo-labels using 4D reconstruction in real-world scale. With our kinematic instruction tuning data for spatio-temporal reasoning, we present ST-VLM, a VLM enhanced for spatio-temporal reasoning, which exhibits outstanding performance on STKit-Bench. Furthermore, we show that ST-VLM generalizes robustly across diverse domains and tasks, outperforming baselines on other spatio-temporal benchmarks (eg, ActivityNet, TVQA+). Finally, by integrating learned spatio-temporal reasoning with existing abilities, ST-VLM enables complex multi-step reasoning. Project page: https://ikodoh.github.io/ST-VLM.

Related papers

• Building a Mind Palace: Structuring Environment-Grounded Semantic Graphs for Effective Long Video Analysis with LLMs [66.57518905079262]
  VideoMind organizes critical video moments into aologically structured semantic graph.<n>"Mind Palace" organizes key information through (i) hand-object tracking, (ii) clustered zones activity representing specific areas of recurring activities, and (iii) environment layout mapping.
  arXiv  Detail & Related papers  (2025-01-08T08:15:29Z)
• Coarse Correspondences Boost Spatial-Temporal Reasoning in Multimodal Language Model [51.83436609094658]
  We introduce Coarse Correspondences, a simple lightweight method that enhances MLLMs' spatial-temporal reasoning with 2D images as input. Our method uses a lightweight tracking model to identify primary object correspondences between frames in a video or across different image viewpoints. We demonstrate that this simple training-free approach brings substantial gains to GPT4-V/O consistently on four benchmarks.
  arXiv  Detail & Related papers  (2024-08-01T17:57:12Z)
• DO3D: Self-supervised Learning of Decomposed Object-aware 3D Motion and Depth from Monocular Videos [76.01906393673897]
  We propose a self-supervised method to jointly learn 3D motion and depth from monocular videos. Our system contains a depth estimation module to predict depth, and a new decomposed object-wise 3D motion (DO3D) estimation module to predict ego-motion and 3D object motion. Our model delivers superior performance in all evaluated settings.
  arXiv  Detail & Related papers  (2024-03-09T12:22:46Z)
• SpOT: Spatiotemporal Modeling for 3D Object Tracking [68.12017780034044]
  3D multi-object tracking aims to consistently identify all mobile time. Current 3D tracking methods rely on abstracted information and limited history. We develop a holistic representation of scenes that leverage both spatial and temporal information.
  arXiv  Detail & Related papers  (2022-07-12T21:45:49Z)
• Exploring Optical-Flow-Guided Motion and Detection-Based Appearance for Temporal Sentence Grounding [61.57847727651068]
  Temporal sentence grounding aims to localize a target segment in an untrimmed video semantically according to a given sentence query. Most previous works focus on learning frame-level features of each whole frame in the entire video, and directly match them with the textual information. We propose a novel Motion- and Appearance-guided 3D Semantic Reasoning Network (MA3SRN), which incorporates optical-flow-guided motion-aware, detection-based appearance-aware, and 3D-aware object-level features.
  arXiv  Detail & Related papers  (2022-03-06T13:57:09Z)
• Spatio-Temporal Self-Attention Network for Video Saliency Prediction [13.873682190242365]
  3D convolutional neural networks have achieved promising results for video tasks in computer vision. We propose a novel Spatio-Temporal Self-Temporal Self-Attention 3 Network (STSANet) for video saliency prediction.
  arXiv  Detail & Related papers  (2021-08-24T12:52:47Z)
• Monocular Quasi-Dense 3D Object Tracking [99.51683944057191]
  A reliable and accurate 3D tracking framework is essential for predicting future locations of surrounding objects and planning the observer's actions in numerous applications such as autonomous driving. We propose a framework that can effectively associate moving objects over time and estimate their full 3D bounding box information from a sequence of 2D images captured on a moving platform.
  arXiv  Detail & Related papers  (2021-03-12T15:30:02Z)
• Tracking from Patterns: Learning Corresponding Patterns in Point Clouds for 3D Object Tracking [34.40019455462043]
  We propose to learn 3D object correspondences from temporal point cloud data and infer the motion information from correspondence patterns. Our method exceeds the existing 3D tracking methods on both the KITTI and larger scale Nuscenes dataset.
  arXiv  Detail & Related papers  (2020-10-20T06:07:20Z)
• Comparison of Spatiotemporal Networks for Learning Video Related Tasks [0.0]
  Many methods for learning from sequences involve temporally processing 2D CNN features from the individual frames or directly utilizing 3D convolutions within high-performing 2D CNN architectures. This work constructs an MNIST-based video dataset with parameters controlling relevant facets of common video-related tasks: classification, ordering, and speed estimation. Models trained on this dataset are shown to differ in key ways depending on the task and their use of 2D convolutions, 3D convolutions, or convolutional LSTMs.
  arXiv  Detail & Related papers  (2020-09-15T19:57:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.