arXiv Computer Vision

BEAST3D: Animal behavioral analysis and neural encoding from multi-view video via Gaussian splatting

BEAST3D is a self‑supervised framework that learns 3D visual representations from unlabeled, calibrated multi‑view video of animals. It uses a vision transformer to predict 3D Gaussian splats, reconstructing held‑out views via differentiable rendering while simultaneously segmenting the animal. The method works with as few as four views, produces viewpoint‑invariant features, and transfers to tasks such as novel view synthesis, multi‑view pose estimation, and neural encoding across four species.

arXiv Computer Vision
Sep 30

Sparse-View Interpretable 3D Animal Behavior Representations for Neural Encoding and Decoding

arXiv:2609.36217v1 Announce Type: new Abstract: A deeper understanding of brain function requires a precise, structured characterization of behavior.Yet, extracting behavioral representations from vi...

By Xinming Dai, Qihang Jin, Tianshu Tan, Baiyuan Chen, Hanrui Lyu, Lenny Aharon, Kyle Daruwalla, Xun Helen Hou, Matthew R. Whiteway, Liam Paninski, Yizi Zhang
arXiv Computer Vision
Sep 30

ORMA: Optimization-based Monocular 4D Reconstruction of Articulated Animals

ORMA is a training‑free framework that reconstructs articulated 4D representations of animals from monocular videos by decoupling pose and shape. It uses predicted pose as a reference for optimization and generative 3D priors to refine shape, aligning the result with the SMAL+ parametric model. The method combines per‑frame pose estimates with globally consistent camera poses, and further refines the reconstruction using self‑supervised DINO correspondences and temporal consistency, achieving improved accuracy on the new PAW4D benchmark and diverse real‑world videos.

By Xuyi Hu, Francesco Palandra, Shangzhe Wu, Daniel Cremers, Riccardo Marin, Silvia Zuffi
arXiv AI
3d ago

Multi-View 3D Reconstruction Using V-JEPA Features

The paper investigates supervised RGB‑only multi‑view 3D reconstruction by integrating video‑pretrained features with camera estimation, scene aggregation, epipolar transport, and residual pointmaps. Experiments on CO3D and indoor datasets compare global versus spatial camera regression and a joint relative‑objective configuration, reporting improvements in trajectory and intrinsics errors but mixed effects on depth accuracy and raw Chamfer distance. The results emphasize that representation adaptation, shape alignment, and raw reconstruction must be evaluated separately before attributing gains to predictive video pretraining.

By Anwar Shamim, Jeffrin Sam
arXiv Computer Vision
Aug 27

PIVOT: A Multi-Trajectory Dataset and Testbed for Pose, Intrinsics, and Novel Viewpoint Evaluation in Real-World 3D Reconstruction

PIVOT is a new multi‑trajectory dataset and evaluation framework that captures real‑world scenes with diverse camera paths, preserving both sensor‑derived measured poses and COLMAP‑optimized poses along with calibrated and optimized intrinsics. It defines three benchmark families—seen vs. unseen trajectory generalization, measured vs. optimized pose sensitivity, and calibrated vs. optimized intrinsics sensitivity—and introduces a directed pose‑space Chamfer distance to assess pose coverage. The first version of PIVOT includes five scenes recorded with a DJI Mini 4 Pro and offers an open processing and Nerfstudio‑based evaluation toolchain, revealing a consistent quality gap between held‑out and unseen trajectories and significant sensitivity to pose source and camera intrinsics.

By Mary Raymond
Hugging Face Trending Papers
6d ago

Multi-View 3D Reconstruction Using V-JEPA Features

The paper investigates supervised RGB‑only multi‑view 3D reconstruction by leveraging video‑pretrained features and a pipeline that includes camera estimation, scene aggregation, epipolar transport, and residual pointmaps. Experiments on CO3D and indoor datasets compare global versus spatial camera regression and a joint relative‑objective configuration, revealing trade‑offs in trajectory, depth, and Chamfer distance metrics. The results emphasize that improvements in reconstruction stem from specific representation adaptations and component interactions rather than solely from predictive video pretraining.

arXiv Computer Vision
Sep 7

Object Concepts Emerge from Motion

The paper introduces a biologically inspired framework that learns object‑centric visual representations from raw videos without human annotations or camera calibration. By using motion boundaries detected via optical flow and clustering to create pseudo‑instance masks, the method supervises a single‑image encoder with pixel‑level pairwise metric learning. Training on 195 million pseudo‑labeled frames and expanding to 421 million frames through Motion‑Verified Self‑Training, the approach yields Swin‑based encoders that outperform or match supervised and self‑supervised baselines on tasks such as monocular depth estimation, 3D object detection, 3D occupancy prediction, and end‑to‑end planning.

By Boshi Li, Xiaohui Wang, Xiaoyang Wu, Zhichao Li, Ya Yang, Naiyan Wang
arXiv Computer Vision
Sep 3

Kirin: Animal Motion Generation from In-the-Wild Video

Kirin is a new framework that reconstructs 3D animal motion from in‑the‑wild videos, learns motion priors at scale, and generates realistic motion conditioned on text and image. It introduces AiM3D, the first large‑scale dataset of aligned video‑text‑motion tuples for quadruped animals, and uses an off‑the‑shelf image‑to‑3D model to automatically rig and animate 3D meshes with the generated motion. The framework and dataset provide a foundation for large‑scale, text and image‑conditioned animal motion generation and animation.

By Brian Nlong Zhao, Zhuoyang Pan, James M. Rehg, Jiajun Wu, Shangzhe Wu
arXiv AI
Oct 9

SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild

SAM 3D Animal is a promptable framework that performs multi‑animal 3D reconstruction from a single image, using the SMAL+ parametric animal model. It jointly reconstructs multiple instances and allows flexible prompts such as keypoints and masks to better handle crowded and occluded scenes. The authors also introduce Herd3D, a new dataset of over 5,000 images that increases species, interaction, and occlusion diversity, and demonstrate state‑of‑the‑art performance on several benchmark datasets.

By Xuyi Hu, Jin Lyu, Jiuming Liu, Yebin Liu, Silvia Zuffi, Liang An, Stefan Goetz
Hugging Face Trending Papers
Oct 29

LucidLyra: Feed-Forward Generative 3D Scene Reconstruction and Joint Understanding via Multi-Task Distillation

LucidLyra is a framework that reconstructs 3D scenes from video latents in a feed‑forward manner while simultaneously providing open‑vocabulary semantics and instance cues. By distilling knowledge from a camera‑controlled video diffusion model and frozen 2D vision foundation models, it jointly infers 3D Gaussians that encode geometry, appearance, and semantic information without requiring per‑scene optimization or manual 3D annotations. The resulting scenes support novel‑view rendering, open‑vocabulary segmentation, and object grouping, and can be extended to monocular video for dynamic 3D reconstruction.

Hugging Face Trending Papers
Aug 11

CasDeblurGS: Cascaded 2D-to-3D Multi-View Consistency for 3D Gaussian Splatting from Two Blurry Images

Free-viewpoint 3D scene media is increasingly important for immersive applications, yet practical capture often suffers from severe view sparsity and motion blur. Although neural rendering has advanced sparse-view synthesis, existing blur-aware methods typically require substantial multi-view redundancy, accurate camera poses, or costly per-scene optimization.