All papers

Introduces a test-time verification method using latent world models to detect critical states and select better actions in flow-based policies.

Proposes an active visual tracking framework that optimizes robot motion to prevent pedestrian occlusions when following a target person.

Analyzes the planning-execution gap in frozen world-model planners and adapts trajectory representations to downstream controller shifts.

Develops a control barrier function allocator to preserve wrench feasibility during slow servo reconfigurations on tiltrotor aerial robots.

Designs an onboard pipeline using a lightweight language model to infer obstacle-specific safety clearance for aerial robot navigation around humans.

Presents a closed-form triangle splatting framework for efficient novel-view synthesis by linearly solving spherical harmonics over a known mesh.

Introduces a residual adapter for pretrained general motion trackers to give humanoid robots compliant, force-aware responses to physical disturbances.

This paper introduces an adaptive visual token compression method to reduce latency and computation during vision-language-action robot policy inference.

This paper integrates whole-body model predictive control with online parameter estimation to improve humanoid robot teleoperation under heavy payloads.

This paper presents a method for flow matching robotic policies that guides behavior mode selection through targeted source-space distributions.

This paper presents an autonomous motion planning and vision-guided control framework for fuel-efficient hopping maneuvers of lunar drones.

This paper introduces a framework that learns fixed-size latent temporal representations to guide action policies in long-horizon robotic manipulation.

This paper presents an active inspection planning framework using image-based autoencoder variance to account for harsh lighting variations in orbit.

This paper proposes a reinforcement learning framework trained entirely in simulation for path-safe hydroponic lettuce harvesting with physical robot arms.

This paper introduces a vision-tactile-action model pretrained on human hand data for contact-rich dexterous robotic manipulation.

This paper presents a framework using cached value predictions as memory to guide online reinforcement learning for long-horizon robotic manipulation.

This paper proposes a tokenized joint source-network coding framework to improve video streaming resilience against network packet loss.

This paper presents a control and communication framework using reachable sets and barrier functions to guarantee multi-agent safety during intermittent communication.

This paper introduces a posetal game framework for auditable multi-agent motion planning under partially ordered legal constraints.

This paper establishes analytical geometric bounds on coverage probability and timing uncertainty for multi-agent drone meshes intercepting approaching objects.

This paper investigates internal geometric representations of task-relevant physical concepts within vision-language-action policies for robotic manipulation.

The paper presents a GPU-accelerated robust adaptive control framework that accounts for guaranteed future parameter learning during real-time robotic trajectory planning.

The paper presents a predictive verifier that decides how many actions from a vision-language-action chunk to execute safely.

Applies persistent homology to encode global structural connectivity into learned representations of mechanical metamaterials.

Presents a benchmark to evaluate 3D consistency, input preservation, and content persistence in interactive video world models.

Introduces local appearance purification to filter alignment targets and improve representation learning during diffusion model training.

Provides region-specific reward modeling to fix incorrect wording and glyph errors in visual text generation models.

Combines state-tracking rules with action-conditioned video generation to enable persistent reciprocal interactions in generative gameplay.

Uses a multi-codebook diffusion language model to jointly represent text and discrete visual tokens for multimodal generation.

Introduces an autoregressive foundation model to generate vectorized road networks and building polygons directly from spatial inputs.

Builds a solar-powered edge vision system to detect and quantify insect defoliation severity in grapevine canopies.

Selects maximally informative pixel tokens directly from vision transformers to train segmentation heads under tight annotation budgets.

Learns robot manipulation priors from egocentric human videos using action prediction rather than rule-based hand retargeting.

Prunes Gaussian occupancy models without preset budgets by identifying and removing Gaussians that do not alter voxel classification decisions.

Adds line-feature tracking to GPU-accelerated visual odometry to improve warehouse robot localization in low-texture environments.

Uses synthetic physics-based satellite observations to probe and explain physical feature sensitivity inside geo-foundation model representations.

Evaluates on-policy distillation versus supervised fine-tuning in vision-language models, analyzing how teacher reliability impacts student behavior.

Decomposes agentic spatial reasoning by using a metric 3D perception tool to resolve visual inconsistencies before spatial reasoning.

Uses sparse autoencoders to select diverse training samples and prevent underrepresenting tail concepts across vision and multimodal models.

Combines visual tracking and vocal fingerprinting to identify and monitor individual dogs across real-world multi-dog video recordings.

Estimates 3D robot pose from monocular images using graph-guided 2D keypoints and conditional flow matching.

Presents a dataset of 500 robotically imaged cloth materials for training and evaluating neural reflectance reconstruction models.

Tests whether jointly training action-conditioned world models and robot policy models in a single architecture improves both tasks.

Introduces a Shapley-based attribution method that jointly explains spatial and spectral channel importance in multispectral imagery.

Separates static and dynamic 3D Gaussians to enable efficient real-time volumetric video streaming on mobile devices.

Accelerates diffusion transformer video generation by increasing attention sparsity and reusing search indices across denoising steps.

Develops a domain adaptation framework to improve pedestrian trajectory prediction under varied social couplings and scene shifts.

Introduces a counterfactual benchmark to evaluate whether multimodal models erroneously defer to misleading on-screen text over video evidence.

Presents an unsupervised deep multiview learning framework to correct systematic multitemporal surface reflectance inconsistencies across satellite observations.

Distills multi-view visual cues into a latent flow matching network to improve point cloud upsampling at inference time.

Distills high-fidelity priors into heavily compressed latent spaces to enable faster high-resolution image generation.

Reuses intermediate transformer features across video frames to speed up 3D point cloud semantic segmentation without extra training.

Proposes a local optimization method using Delaunay triangulation to balance geometric consistency with spatial distribution in RANSAC estimation.

Introduces a climate-aware satellite image dataset to evaluate and improve domain adaptive land-cover semantic segmentation across different regions.

Improves on-policy distillation for diffusion models by restricting feature alignment supervision to deep layers and early denoising steps.

Studies how vision-language-action models fail to follow process instructions and introduces a benchmark to measure instruction compliance.

Introduces a robotic system that converts language instructions and visual inputs into physical light projection actions in spatial environments.

Presents a training-free state-correction method to eliminate object residuals during object removal in few-step diffusion models.

Introduces a feed-forward framework with panoramic attention to recover geometrically consistent 3D point clouds from a single panorama.

Improves dynamic 3D Gaussian splatting for fast-moving objects by injecting seeds in under-reconstructed dynamic areas.