Domain adaptation remains a central challenge in 3D vision, especially for multimodal foundation models that align 3D point clouds with visual and textual data. While these models demonstrate strong general capabilities, adapting them to downstream domains with limited data often leads to overfitting and catastrophic forgetting.
arXiv:2607. 07179v1 Announce Type: cross Abstract: Document Visual Question Answering (DocVQA) presents a complex multimodal challenge, requiring models to exploit visual, textual, and layout information from documents.
arXiv:2604. 02546v3 Announce Type: replace-cross Abstract: Pretraining 3D encoders through alignment with Contrastive Language-Image Pre-training (CLIP) has emerged as a promising direction for learning generalizable representations for 3D scene understanding.
Document Visual Question Answering (DocVQA) presents a complex multimodal challenge, requiring models to exploit visual, textual, and layout information from documents. Although Vision-Language Models (VLMs) have shown remarkable performance in text-vision tasks, their robustness and transferability to different document domains remains underexplored.
arXiv:2606. 31148v1 Announce Type: cross Abstract: 3D Visual Grounding (3DVG) aims to localize target objects in 3D scenes given natural language descriptions.
arXiv:2603. 09493v2 Announce Type: replace-cross Abstract: The adaptation of large-scale vision-language models (VLMs) to downstream tasks with limited labeled data remains a significant challenge.
ScenePilot introduces a retrieval‑augmented Grow‑and‑Repair framework for text‑driven 3D indoor scene generation. It uses a Hierarchical Retrieval‑Augmented Planning module to fetch room, group, and anchor layout priors, then incrementally inserts object groups with a base generator, while a Reinforcement Multimodal Repair module performs lightweight local corrections after each insertion and a final global repair. The approach is trained on a new SceneReverse‑17k dataset of perturbed scenes, enabling the policy to predict structured move‑rotate‑scale actions from rendered views, scene state, retrieved priors, and edit history, thereby improving physical plausibility, functional coherence, and controllability without heavy full‑scene optimization.
GoDeep is an annotation‑free method for open‑vocabulary 3D scene understanding that uses a vision‑language model solely as a translator to generate structured, entity‑level descriptions of each image. These descriptions are projected and aggregated in a language‑only embedding space, eliminating the need for a 3D training corpus or domain‑specific encoder. The approach achieves competitive performance on ScanNet++ and a cultural heritage benchmark, accurately localizes out‑of‑vocabulary objects, and offers explainable, point‑level predictions.
arXiv:2606. 05058v1 Announce Type: cross Abstract: Computer-Aided Design (CAD) underpins modern engineering and manufacturing by enabling the creation of precise, editable 3D models.
The paper introduces T23D-CompBench, a new benchmark for fine‑grained text‑to‑3D quality assessment that includes 3,600 textured meshes generated from ten state‑of‑the‑art models and 129,600 human ratings. It also proposes Rank2Score, a two‑stage rank‑learning metric that first trains with supervised contrastive regression and curriculum learning, then refines predictions using mean opinion scores to better align with human judgments. Experiments show Rank2Score outperforms existing metrics and can be used as a reward function for generative model optimization.
The paper introduces a framework for instruction‑guided 3D editing that does not require paired 3D supervision. It distills visual, semantic, and geometric knowledge from foundation models into a 3D editing model using a differentiable rendering pipeline, guided by a 2D visual prior from an image editing model and a semantic prior from a Vision‑Language Model. A 3D‑aware Distribution Matching regularization is added to prevent geometric collapse and ensure realistic 3D outputs, leading to superior instruction fidelity and cross‑view consistency compared to state‑of‑the‑art baselines.
The paper introduces Vision-Free Adaptation (VFA), a method that separates multilingual language enhancement from visual alignment in multimodal large language models. VFA fine‑tunes a base LLM on multilingual text to create a multilingual task vector, which is then merged with the vision‑aligned task vector of an existing MLLM. Experiments on five MLLMs and six multilingual benchmarks show consistent gains while preserving multimodal and text‑only performance, and using less than 2% of text data narrows the performance gap to fully multimodal‑trained models.
arXiv:2511.14086v2 Announce Type: replace-cross Abstract: Despite recent progress in 3D-LLMs, they remain limited in accurately grounding language to visual and spatial elements in 3D environments. T...