Cambrian-1 introduces a vision-centric multimodal large language model (MLLM) leveraging 20+ vision encoders and a novel Spatial Vision Aggregator (SVA), achieving state-of-the-art performance.

Shengbang Tong, Ellis Brown, Penghao Wu et al.

VGA reduces hallucinations in GUI understanding via image-centric fine-tuning using a 63.8k dataset and FAC method.

Ziyang Meng, Yu Dai, Zezheng Gong et al.

HiFiAlbedo estimates high-fidelity facial albedo from a single image via texture quantization.

Zimin Ran, Xingyu Ren, Xiang An et al.

Joint channel, layer, and block pruning via MINLP with accurate latency modeling achieves high pruning ratios and improved speed/accuracy trade-offs.

Xinglong Sun, Barath Lakshmanan, Maying Shen et al.

Proposes diffusion-based autoregressive image generation without vector quantization, achieving FID<3.5 on ImageNet 256×256.

Tianhong Li, Yonglong Tian, He Li et al.

Introduces GeoGPT4V, leveraging GPT-4 and GPT-4V to generate aligned geometric problems and images, significantly enhancing multi-modal geometric reasoning.

Shihao Cai, Keqin Bao, Hangyu Guo et al.

CoMM dataset employs multi-perspective filtering with Llama3 and CLIP, significantly improving content coherence and alignment, boosting multimodal model performance.

Wei Chen, Lin Li, Yongqi Yang et al.

L4GM is a fast 4D Gaussian-based model that generates animated 3D objects from a single-view video in one second, outperforming traditional methods in speed and quality.

Jiawei Ren, Kevin Xie, Ashkan Mirzaei et al.

Glyph-ByT5-v2 leverages large-scale multilingual datasets and preference learning to achieve high-accuracy multilingual visual text rendering with enhanced aesthetics.

Zeyu Liu, Weicong Liang, Yiming Zhao et al.

Introduces Med-HallMark benchmark and MediHall Score for hierarchical, multi-task hallucination detection in medical LVLMs.

Jiawei Chen, Dingkang Yang, Tong Wu et al.

CarLLaVA uses LLaVA vision encoder and LLaMA backbone, achieving state-of-the-art closed-loop driving with only camera input, outperforming previous methods by 458%.

Katrin Renz, Long Chen, Ana-Maria Marcu et al.

D-NPC introduces dynamic neural point clouds for monocular scene synthesis, achieving real-time high-quality novel views.

Moritz Kappel, Florian Hahlbohm, Timon Scholz et al.

Introduces ImageNet3D, a large-scale dataset with 200 categories, 86,000+ instances annotated with 6D pose, position, and natural language descriptions, advancing general-purpose 3D understanding.

Wufei Ma, Guanning Zeng, Guofeng Zhang et al.

HOT3D dataset offers multi-view images and annotations for 3D hand and object tracking, boosting research.

Prithviraj Banerjee, Sindi Shkodrani, Pierre Moulon et al.

Analyzed 60,000 fine-tuned diffusion models' weight space; introduced weights2weights subspace for sampling, editing, and inversion tasks.

Amil Dravid, Yossi Gandelsman, Kuan-Chieh Wang et al.

Depth Anything V2 leverages synthetic images, larger teacher models, and large-scale pseudo-labels to achieve finer, more robust monocular depth estimation, over 10x faster than prior models.

Lihe Yang, Bingyi Kang, Zilong Huang et al.

Proposes Sketchpad, enabling multimodal models to draw visual sketches, boosting math and visual reasoning by over 12%.

Yushi Hu, Weijia Shi, Xingyu Fu et al.

Yo’LLaVA employs learnable latent tokens for few-shot personalization, significantly enhancing personalized vision-language understanding.

Thao Nguyen, Haotian Liu, Yuheng Li et al.

OmniTokenizer is a unified image-video tokenizer using decoupled spatial-temporal transformer architecture, achieving SOTA reconstruction with 1.11 FID on ImageNet.

Junke Wang, Yi Jiang, Zehuan Yuan et al.

LVNet introduces a hierarchical keyframe selector (HKS) for efficient long-video QA, achieving 61.1% accuracy with only 12 frames, without training.

Jongwoo Park, Kanchana Ranasinghe, Kumara Kahatapitiya et al.