MotionBits: Video Segmentation through Motion-Level Analysis of Rigid Bodies

Howard H. Qian; Kejia Ren; Yu Xiang; Vicente Ordonez; Kaiyu Hang

cs.CVMar 6, 2026

MotionBits: Video Segmentation through Motion-Level Analysis of Rigid Bodies

Authors: Howard H. Qian, Kejia Ren, Yu Xiang, Vicente Ordonez, Kaiyu Hang

Organizations: Department of Computer Science Rice University · Department of Computer Science University of Texas at Dallas

Abstract

Rigid bodies constitute the smallest manipulable elements in the real world, and understanding how they physically interact is fundamental to embodied reasoning and robotic manipulation. Thus, accurate detection, segmentation, and tracking of moving rigid bodies is essential for enabling reasoning modules to interpret and act in diverse environments. However, current segmentation models trained on semantic grouping are limited in their ability to provide meaningful interaction-level cues for completing embodied tasks. To address this gap, we introduce MotionBit, a novel concept that, unlike prior formulations, defines the smallest unit in motion-based segmentation through kinematic spatial twist equivalence, independent of semantics. In this paper, we contribute (1) the MotionBit concept and definition, (2) a hand-labeled benchmark, called MoRiBo, for evaluating moving rigid-body segmentation across robotic manipulation and human-in-the-wild videos, and (3) a learning-free graph-based MotionBits segmentation method that outperforms state-of-the-art embodied perception methods by 37.3% in macro-averaged mIoU on the MoRiBo benchmark. Finally, we demonstrate the effectiveness of MotionBits segmentation for downstream embodied reasoning and manipulation tasks, highlighting its importance as a fundamental primitive for understanding physical interactions.

Figures & tables

Figure 1: The top row shows a robot physically interacting with a variety of complex composite objects, each constructed from colored blocks that have been glued together. The bottom row highlights the output of a standard semantic segmentation model, which incorrectly over-segments the objects. In contrast, the proposed MotionBits segmentation correctly groups composite objects. Example real-world robotic applications requiring accurate MotionBits segmentation are shown in Fig. 6 .

Figure 2: Illustration of spatial twist for two rigid objects. Transparent objects represent initial positions, and solid objects represent positions after motion. The translations of body frames are represented by purple linear velocity vectors $\upsilon_{\{x\}}$ . Although $\upsilon_{\{a_{1}\}}\neq\upsilon_{\{a_{2}\}}$ and $\upsilon_{\{b_{1}\}}\neq\upsilon_{\{b_{2}\}}$ , transforming them to the fixed world frame $\{s\}$ yields identical linear velocities $\upsilon_{\{s\}}^{x}$ for frames on the same rigid object $x$ but distinct linear velocities for different rigid objects. Intuitively, the instantaneous motion observed at the world frame will appear identical for body frames on the same rigid object, regardless of their local motions, but will differ across rigid bodies.

Track	#Scenes	Frames/Scene	Masks/Scene	Source Dataset	Actions and Motions
Robotic Manipulation	270	$15.9\pm 0.7$	$5.2\pm 1.5$	BridgeData V2 [ 35 ]	Push, Grasp, Pick-&-Place
Human-in-the-Wild	79	$11.0\pm 0.0$	$10.3\pm 8.6$	SA-V [ 2 ]	Human-Object Interactions

Table 1: Statistics of the MoRiBo benchmark for evaluating moving rigid-body segmentation. Each video has resolution of 480

\times

640 or 640

\times

480, and hand-annotated segmentation masks are provided for the last frame of each video.

Figure 3: Examples from the new MoRiBo benchmark across both tracks, robotic manipulation and human-in-the-wild. Hand-labeled final-frame segmentation masks of moving rigid bodies are provided for every video.

Figure 5

Figure 4: Our learning-free graph-based method for online MotionBits segmentation.

		Overlap				Boundary
	Method	P	R	$\text{F}_{1}$	mIoU	P	R	$\text{F}_{1}$
Robotic Manipulation	VideoLLaMa [ 26 ]	47.2	9.2	8.6	7.4	43.8	8.4	7.7
	InternVideo [ 27 ]	49.5	17.7	15.0	13.9	44.2	14.2	13.8
	QwenVL [ 25 ]	19.1	28.7	24.6	21.5	17.7	25.1	21.9
	SAMIV [ 8 ]	69.5	31.0	27.0	24.3	61.6	27.2	25.8
	OCLR Flow [ 9 ]	52.5	16.9	24.4	10.2	23.0	14.3	17.8

Table 2: Quantitative results on the MoRiBo benchmark (%). Our method surpasses alternatives across both the Robotic Manipulation and Human-in-the-Wild tracks.

Figure 6: Application examples of MotionBits for downstream tasks in a tabletop scene where a robot autonomously interacts with complex, composite objects. QwenVL [ 25 ] fails to produce an accurate response in the VQA task. However, using our MotionBits segmentation as overlaid marks [ 14 ] improves VQA response accuracy. Comparing object segmentation among SAMIV [ 8 ] , SAM [ 6 ] , and our MotionBits method shows that incorporating rigid-body motions as segmentation cues gives more accurate object-level insight. For the tower stacking task, over-segmentation of the composite objects by SAM [ 6 ] generates robot actions that result in grasp and stack failures. In contrast, our method’s MotionBits segmentation enables successful grasping and stacking.

Method	# Stacked Objs	SR	Description of Failures
SAM [ 6 ]	4	0 / 10	Grasp & Stack Failures
SAMIV [ 8 ]	0	0 / 10	No Segmentation
QwenVL [ 25 ]	2	0 / 10	Poor/No Segmentation
Ours	37	6 / 10	Cluttered Grasp Failures

Table 3: Results of the tower stacking task guided by object segmentation masks.

Appendix figures & tables8 assets

Supplementary material from the paper’s appendix.

Appendix

Figure S7: Iterative soft label propagation visualized ( Section 5.2 and line 8 of Algorithm 1 ). First, a set of seed nodes $S$ is uniformly selected from the set of graph nodes $\mathcal{M}$ , and each seed node $s\in S$ is assigned a different “colored paint”. Then, by continuously pouring the assigned color of paint onto each seed node and allowing the paints to flow and blend naturally according to the structure of graph $G$ , we capture spatial twist motion regions through a color-coded soft MotionBits segmentation mask.

Figure S8: Examples of Qwen2.5VL-32B [ 25 ] responses to different prompts. The prompt that empirically yielded the most accurate moving rigid-body segmentation results (highlighted) was used to evaluate VLMs on the MoRiBo benchmark, reflected in Section 6.1 .

Figure S9: Additional example of MotionBits segmentation applied to the downstream task of opening a cabinet drawer. When a person interacts with the drawer, our method accurately segments the relevant MotionBits, including the drawer door and the two components of the drawer slider. This segmentation provides interaction-level cues that enable an embodied system to reason about and manipulate the cabinet.

Figure S10: Additional example of MotionBits segmentation applied to the downstream task of manipulating the water bottle handle. When a person manipulates the water bottle handle, our method accurately segments the handle as a MotionBit, which provides interaction-level cues that enable a robot to manipulate the handle in a similar manner.

Figure S11: Extended qualitative comparisons of moving rigid-body segmentation on the Robotic Manipulation track of the MoRiBo benchmark between VideoLLaMa [ 26 ] , InternVideo [ 27 ] , QwenVL [ 25 ] , SAMIV [ 8 ] , OCLR TTA [ 9 ] and our method. These results demonstrate that existing perception methods lack fundamental motion-level grounding, which is essential for embodied systems to reason about and interact with real-world environments. Our method is able to effectively identify moving rigid bodies within a scene.

Figure S12: Extended qualitative comparisons of moving rigid-body segmentation on the Human-in-the-Wild track of the MoRiBo benchmark between VideoLLaMa [ 26 ] , InternVideo [ 27 ] , QwenVL [ 25 ] , SAMIV [ 8 ] , OCLR TTA [ 9 ] and our method. These results demonstrate that existing perception methods lack fundamental motion-level grounding, which is essential for embodied systems to reason about and interact with real-world environments. Our method is able to effectively identify moving rigid bodies within a scene.

Figure S13: More examples from the Robotic Manipulation track of the proposed MoRiBo benchmark ( Section 4 ).

Figure S14: More examples from the Human-in-the-Wild track of the proposed MoRiBo benchmark ( Section 4 ).

Explore similar work

GenMatter: Perceiving Physical Objects with Generative Matter Models
Apr 24, 2026Eric Li, Arijit Dasgupta, Yoni Friedman +5Video Object Segmentation Object-Centric Representation Learning
1MIT CSAIL · 2MIT BCS
$Δ$ ynamics: Language-Based Representation for Inferring Rigid-Body Dynamics From Videos
May 20, 2026Chia-Hsiang Kao, Cong Phuoc Huynh, Chien-Yi Wang +5VLM Reasoning Physical Reasoning in Video
1Cornell University · 2Amazon
Towards Motion-aware Referring Image Segmentation
Mar 18, 2026Chaeyun Kim, Seunghoon Yi, Yejin Kim +2Referring Image Segmentation Multimodal Contrastive Learning
Seoul National University · AIM Intelligence

Apr 24, 2026J/K move · Enter open · S save

GenMatter: Perceiving Physical Objects with Generative Matter Models

Eric Li, Arijit Dasgupta, Yoni Friedman, Mathieu Huot+4

1MIT CSAIL · 2MIT BCS

Video Object Segmentation Object-Centric Representation Learning Structure from Motion 3D Scene Understanding

Open paper AskPDF arXiv

Human visual perception offers valuable insights for understanding computational principles of motion-based scene interpretation. Humans robustly detect and segment moving entities that constitute independently moveable chunks of matter, whether observing sparse moving dots, textured surfaces, or naturalistic scenes. In contrast, existing computer vision systems lack a unified approach that works across these diverse settings. Inspired by principles of human perception, we propose a generative model that hierarchically groups low-level motion cues and high-level appearance features into particles (small Gaussians representing local matter), and groups particles into clusters capturing coherently and independently moveable physical entities. We develop a hardware-accelerated inference algorithm based on parallelized block Gibbs sampling to recover stable particle motion and groupings. Our model operates on different kinds of inputs (random dots, stylized textures, or naturalistic RGB video), enabling it to work across settings where biological vision succeeds but existing computer vision approaches do not. We validate this unified framework across three domains: on 2D random dot kinematograms, our approach captures human object perception including graded uncertainty across ambiguous conditions; on a Gestalt-inspired dataset of camouflaged rotating objects, our approach recovers correct 3D structure from motion and thereby accurate 2D object segmentation; and on naturalistic RGB videos, our model tracks the moving 3D matter that makes up deforming objects, enabling robust object-level scene understanding. This work thus establishes a general framework for motion-based perception grounded in principles of human vision.

Abstract

Explore similar work

GenMatter: Perceiving Physical Objects with Generative Matter Models

ΔΔΔynamics: Language-Based Representation for Inferring Rigid-Body Dynamics From Videos

Towards Motion-aware Referring Image Segmentation

$Δ$ ynamics: Language-Based Representation for Inferring Rigid-Body Dynamics From Videos