Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Jiebo LuoSuggested from co-authorship
- Jianguo ZhangSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works45 from public data
- Video Understanding With Large Language Models: A Survey332
The emergent capabilities of Vid-LLMs are surprisingly advanced, particularly their ability for open-ended multi-granularity (abstract, temporal, and spatiotemporal) reasoning combined with common-sense knowledge, suggesting a promising path for future video understanding.
- DNAGPT: A Generalized Pre-trained Tool for Multiple DNA Sequence Analysis Tasks92
Evaluation of genomic signal and region recognition, mRNA abundance regression, and artificial genome generation tasks demonstrates DNAGPT’s superior performance compared to existing models designed for specific downstream tasks, benefiting from pre-training using the newly designed model structure.
- NTIRE 2024 Challenge on Image Super-Resolution (×4): Methods and Results83
This paper reviews the NTIRE 2024 challenge on image super-resolution (×4), highlighting the solutions proposed and the outcomes obtained, and offers a comprehensive overview of current trends in this field.
- CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs74
This research extensively investigates the capability of LMMs to perceive fine-grained visual details when dealing with multiple input images, and develops a Contrastive Chain-of-Thought (CoCoT) prompting approach based on multi-input multimodal models to enhance model performance.
- Prototype Knowledge Distillation for Medical Segmentation with Missing Modality74
This paper proposes a prototype knowledge distillation (ProtoKD) method that can not only distillate the pixel-wise knowledge of multi-modal data to single-modality data but also transfer intra-class and inter-class feature variations, such that the student model could learn more robust feature representation from the teacher model and inference with only one single modality data.
- Feature Alignment and Uniformity for Test Time Adaptation74
This work addresses TTA as a feature revision problem due to the domain gap between source domains and target domains, and proposes a memorized spatial local clustering strategy to align the representations among the neighborhood samples for the upcoming batch.
- GPT-4V(ision) as A Social Media Analysis Engine54
GPT-4V shows effectiveness in these tasks, exhibiting capabilities like joint image–text understanding, contextual and cultural awareness, and commonsense knowledge application, however, challenges persist, including struggles with multilingual social multimedia comprehension and difficulty in adapting to the latest social media trends.
- Cross Contrasting Feature Perturbation for Domain Generalization33
Comprehensive experiments show that the proposed CCFP framework outperforms the previous state-of-the-art, and quantitative analyses illustrate that the approach can alleviate the domain shift problem in out-of-distribution (OOD) scenarios.
- On Path to Multimodal Generalist: General-Level and General-Bench29
This project introduces General-Level, an evaluation framework that defines 5-scale levels of MLLM performance and generality, offering a methodology to compare MLLMs and gauge the progress of existing systems towards more robust multimodal generalists and, ultimately, towards AGI.
- Why Reasoning Matters? A Survey of Advancements in Multimodal Reasoning (v1)24
A thorough and up-to-date comparison of reasoning techniques in both textual and multimodal LLMs is offered, highlighting practical methods for post-training optimization and test-time inference.
- Rethinking Alignment and Uniformity in Unsupervised Semantic Segmentation22
Experimental results on multiple semantic segmentation benchmarks show that the proposed Semantic Attention Network (SAN) framework specializes in catching semantic representations, which outperforms all the unpretrained and even several pretrained methods.
- WorldGenBench: A World-Knowledge-Integrated Benchmark for Reasoning-Driven Text-to-Image Generation20
A benchmark designed to systematically evaluate T2I models' world knowledge grounding and implicit inferential capabilities, covering both the humanities and nature domains is introduced, and a structured metric that measures how well generated images satisfy key semantic expectations is proposed.
- UniVA: Universal Video Agent towards Open-Source Next-Generation Video Generalist16
UniVA, an open-source, omni-capable multi-agent framework for next-generation video generalists that unifies video understanding, segmentation, editing, and generation into cohesive workflows, and UniVA-Bench, a benchmark suite of multi-step video tasks spanning understanding, editing, segmentation, and generation, to rigorously evaluate such agentic video systems.
- 16
- 14
- Aggregation of Disentanglement: Reconsidering Domain Variations in Domain Generalization12
A novel paradigm called Domain Disentanglement Network (DDN) is proposed to disentangle the domain expert features from the source domain images and aggregate the sourcedomain expert features for representing the target test domain.
- TransVLAD: Focusing on Locally Aggregated Descriptors for Few-Shot Learning12
This paper is the first to prove self-supervised learning like masked autoencoders (MAE) can deal with the overfitting of transformers in few-shot image classification, and proposes two methods to mitigate few-shot biases, supervision bias and simple-characteristic bias.
- How LLMs React to Industrial Spatio-Temporal Data? Assessing Hallucination with a Novel Traffic Incident Benchmark Dataset9
This study introduces a novel cross-lingual benchmark dataset comprising nearly 99,869 real traffic incident records from Vienna to assess the robustness of state-of-the-art LLMs in the spatial and temporal domains for traffic incident classification and incorporated Retrieval Augmented Generation to further examine LLM hallucinations in both spatial and temporal domains.
- MIPI 2024 Challenge on Demosaic for Hybridevs Camera: Methods and Results9
A total of 110 participants from both industrial and academic backgrounds contributed many valuable solutions to address the difficulty of the restoration of HybridEVS’s raw data, thus raising the reconstructed performance to a new height.
- DeFine: Decision-Making with Analogical Reasoning over Factor Profiles8
This paper introduces \textsc{DeFine}, a modular framework that constructs probabilistic factor profiles from complex scenarios and integrates these profiles with analogical reasoning, leveraging insights from similar past experiences to guide LLMs in making critical decisions in new situations.
- EIBC: a deep learning framework for Chinese toponym recognition with multiple layers7
Experimental results show that the EIBC model outperforms existing some state-of-the-art Chinese toponym recognition models and incorporates CRF for global optimization of toponym sequence labels, enhancing toponym recognition effectiveness.
- GaussianStyle: Gaussian Head Avatar via StyleGAN6
The GaussianStyle preserves structural information, such as expressions and poses, using Gaussian points, while projecting the implicit volumetric representation into Style-GAN to capture high-frequency details and mitigate the over-smoothing commonly observed in neural texture rendering.
- 6
- Self-Boosting Vision-Language Models with Noisy Student On-Policy Self-Distillation5
This work proposes NOPD (Noisy Student On-Policy Self-Distillation), a simple yet effective self-distillation approach that improves VLMs without any external models or ground-truth answers, and demonstrates that NOPD is a general approach to enhance VLMs, achieving improvements across three models on 12 benchmarks.
- VideoWeaver: Evaluating and Evolving Skills for Agentic Long Video Generation5
VideoWeaver is introduced, an agent harness and benchmark that evaluates and evolves skills for long video generation and proposes an evidence-grounded agent-as-judge that inspects both the execution trace and the final video to diagnose process and output failures.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-12. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.