Vision & Video

VideoRAE explores whether frozen video foundation model representations can become compact, reconstructable, and generation-friendly video latents. The work targets a key limitation of conventional 3D-VAEs: latents optimized mainly for pixel reconstruction.

A new arXiv paper studies a multi-expert OCR pipeline for historical Manchu documents with sharply different visual styles. The system routes each page to a specialist model using a lightweight page-level image classifier.

A new arXiv paper on the ABAW11 challenge argues that valence-arousal estimation, expression recognition, and action unit detection should not necessarily share the same fusion strategy.

A new arXiv paper proposes VisionScreen, a vision model that extends the Screening mechanism from language modeling to image recognition. Instead of assigning relative softmax weights to all patches, it aims to explicitly exclude low-relevance visual tokens.

A new arXiv paper proposes modeling choreography as a sequence of interpretable atomic movements rather than only as continuous motion. The approach separates dance planning from motion completion to improve structure, rhythm alignment, and controllability.

A new arXiv paper introduces Peak-End-Net, a lightweight framework for video aesthetic assessment inspired by the psychological peak-end rule. Instead of treating all frames equally, it emphasizes salient moments and the ending of a video.

A new arXiv paper introduces TCA-Net, a low-light image enhancement network centered on more reliable fusion between intensity and chromaticity streams. Its key idea is to replace fixed Top-K sparse attention with thresholded cross-attention.

PiVoT introduces a variational inference approach for real-time multi-object detection and tracking in noisy radar point clouds. It targets clutter-heavy, data-scarce scenarios without relying on external clustering or learned detectors.

A new arXiv paper introduces Spatial Texture-Atlas Splatting, a representation that bakes high-frequency, view-independent texture detail into compact atlases. The authors report up to a fivefold speedup over 3DGS while preserving state-of-the-art visual fidelity and enabling real-time 4K rendering at 60 FPS on consumer hardware.

RainDancer targets video deraining by combining RGB frames with event-camera cues, while avoiding naive cross-modal fusion. Its central idea is to decompose rain and background components before letting the two modalities interact.