7 papers
DeepGen 1.0: A Lightweight Unified Multimodal Model for Advancing Image Generation and Editing
Dianyi Wang, Ruihang Li, Feng Han +17
Current unified multimodal models for image generation and editing typically rely on massive parameter scales (e.g., >10B), entailing prohibitive training costs and deployment foot…
HSD: Training-Free Acceleration for Document Parsing Vision-Language Models with Hierarchical Speculative Decoding
Wenhui Liao, Hongliang Li, Pengyu Xie +15
Document parsing is a fundamental task in multimodal understanding, supporting a wide range of downstream applications such as information extraction and intelligent document analy…
Modality Gap-Driven Subspace Alignment Training Paradigm For Multimodal Large Language Models
Xiaomin Yu, Yi Xin, Yuhui Zhang +12
Despite the success of multimodal contrastive learning in aligning visual and linguistic representations, a persistent geometric anomaly, the Modality Gap, remains: embeddings of d…
UniGenBench++: A Unified Semantic Evaluation Benchmark for Text-to-Image Generation
Yibin Wang, Zhimin Li, Yuhang Zang +8
Recent progress in text-to-image (T2I) generation underscores the importance of reliable benchmarks in evaluating how accurately generated images reflect the semantics of their tex…
UniMedVL: Unifying Medical Multimodal Understanding and Generation through Observation-Knowledge-Analysis
Junzhi Ning, Wei Li, Cheng Tang +24
Medical workflows routinely combine reading images with producing visual and textual outputs, making both image understanding and generation central to medical AI. Most existing sy…
Benchmarking Multimodal Knowledge Conflict for Large Multimodal Models
Yifan Jia, Kailin Jiang, Yuyang Liang +11
Large Multimodal Models(LMMs) face notable challenges when encountering multimodal knowledge conflicts, particularly under retrieval-augmented generation(RAG) frameworks where the…
Sparse-Tuning: Adapting Vision Transformers with Efficient Fine-tuning and Inference
Ting Liu, Xuyang Liu, Liangtao Shi +6
Parameter-efficient fine-tuning (PEFT) has emerged as a popular solution for adapting pre-trained Vision Transformer (ViT) models to downstream applications by updating only a smal…