arXiv:2606.26738v3 Announce Type: replace Abstract: While recent advancements in generative image editing models have achieved stunning visual fidelity, it remains an open question whether these syst...
The paper introduces a prior-guided material reparameterization for physically based inverse rendering, leveraging learned material predictions to restrict the optimization space without altering the photometric objective. By fusing multi-view predictions onto a mesh and constructing a permutohedral lattice in prior feature space, surface points with similar predictions share material variables, which are jointly optimized with geometry and illumination. Experiments on synthetic benchmarks and Stanford-ORB show improved material decomposition and relighting, with PSNR gains of 1.5–2.5 dB over strong baselines and a faster optimization schedule that maintains high relighting quality.
PoseDreamer is a new pipeline that uses diffusion models to generate large‑scale synthetic datasets for 3D human mesh estimation, providing 3D mesh annotations that remain aligned with the generated images. The system incorporates controllable image generation, Direct Preference Optimization for control alignment, curriculum‑based hard sample mining, and multi‑stage quality filtering to produce over 500,000 high‑quality samples with a 76% improvement in image‑quality metrics over traditional rendering‑based datasets. Models trained on PoseDreamer match or surpass those trained on real‑world or conventional synthetic data, and combining PoseDreamer with synthetic datasets yields better performance than mixing real and synthetic data alone.
The paper introduces a benchmark called Shedding Light to evaluate how well generative image models understand and reproduce lighting. The benchmark tests models by asking them to inpaint a simple object, called a light probe, into real photographs and then compares the generated probe to the ground truth to assess lighting direction, colour, and radiance. The authors provide a scalable protocol and open-source code and data for systematic assessment of photometric accuracy in future models.
arXiv:2609.14352v1 Announce Type: new Abstract: AI-generated image detection has attracted increasing attention, but existing evaluations mainly focus on natural images, leaving AI-generated document...
arXiv:2608. 10628v1 Announce Type: cross Abstract: Long-document understanding often requires reasoning over many visually rich pages, making inference costly and prone to context rot.
NOFACE3D is a training‑free framework that refines fine‑scale, detail‑critical regions in 3D assets generated from a single image. It extracts a selected submesh, scales it to the generator’s input volume, and re‑encodes it into a full‑resolution latent grid, refining the region with flow‑based sampling guided by a spatially varying noise schedule. The refined region is then reintegrated, preserving surrounding geometry and appearance, and the method is evaluated on PAniC‑3D, TexVerse, and a new BigDetail benchmark, showing sharper geometry and finer texture detail. whyItMatters: The approach improves the fidelity of small, high‑curvature structures in image‑to‑3D models without requiring additional training, addressing a key limitation in current generators.
DR Pixel introduces a page‑native deep research agent that uses rendered document page images as the sole unit for retrieval and reasoning, bypassing OCR. The open‑source system operates on 1.06 M pages and, without training, outperforms OCR‑based baselines on GPT‑5.4, Gemini‑3.5‑Flash, and Qwen3.6‑27B while requiring fewer tool calls. Training with counterfactual credit reweighting further boosts retrieval recall by ~15 points and accuracy by >10 points on the DRPixelBench benchmark, and brings the model within 1.8 points of an oracle on the ViDoRe‑V3 out‑of‑distribution test.
The paper argues that large-scale text‑to‑image models can be trained effectively on ImageNet, provided the dataset is enriched with carefully crafted text and image augmentations. Using this approach, the authors match the performance of state‑of‑the‑art models such as FLUX, surpassing SD3 on GenEval by +5 points and SDXL on DPGBench by +12, while employing only 1/1000th of the training images and significantly fewer parameters. The method requires just 500 hours of H100 GPU time, making it a more reproducible and accessible alternative to massive web‑scraped datasets.
ShredBench is a new benchmark for evaluating multimodal large language models (MLLMs) on the task of reconstructing documents from shredded fragments. It uses an automated pipeline that generates fragmented documents from Markdown, covering English, Chinese, code, and table scenarios with varying fragmentation levels. Experiments show that while MLLMs perform well on intact documents, their performance drops sharply when documents are shredded, revealing a gap in fine‑grained cross‑modal reasoning.
arXiv:2609.01409v1 Announce Type: new Abstract: Vision-language models (VLMs) have shown strong performance in generating scientific figures from text or images. However, producing publication-ready...
arXiv:2602. 15257v3 Announce Type: replace-cross Abstract: We present the first comprehensive, large-scale study of training long-context vision language models up to 344K context, targeting long-document visual question answering with measured transfer to long-context text.