Zhihang Fu
YOU?
Author Swipe
View article: Human-centered Interactive Learning via MLLMs for Text-to-Image Person Re-identification
Human-centered Interactive Learning via MLLMs for Text-to-Image Person Re-identification Open
Despite remarkable advancements in text-to-image person re-identification (TIReID) facilitated by the breakthrough of cross-modal embedding models, existing methods often struggle to distinguish challenging candidate images due to intrinsi…
View article: A Study on the User Group Classification of Online Q&A Communities Based on Knowledge Sharing Behavior
A Study on the User Group Classification of Online Q&A Communities Based on Knowledge Sharing Behavior Open
User classification is crucial for studying knowledge-sharing behavior in online communities. A well-designed classification system can reveal the distinct characteristics and needs of different user groups, enhance the specificity of info…
View article: Structure-aware Domain Knowledge Injection for Large Language Models
Structure-aware Domain Knowledge Injection for Large Language Models Open
This paper introduces a pioneering methodology, termed StructTuning, to efficiently transform foundation Large Language Models (LLMs) into domain specialists. It significantly reduces the training corpus needs to a mere 5% while achieving …
View article: Optimal Parameter and Neuron Pruning for Out-of-Distribution Detection
Optimal Parameter and Neuron Pruning for Out-of-Distribution Detection Open
For a machine learning model deployed in real world scenarios, the ability of detecting out-of-distribution (OOD) samples is indispensable and challenging. Most existing OOD detection methods focused on exploring advanced training skills o…
View article: Uncertainty-aware Unsupervised Multi-Object Tracking
Uncertainty-aware Unsupervised Multi-Object Tracking Open
Without manually annotated identities, unsupervised multi-object trackers are inferior to learning reliable feature embeddings. It causes the similarity-based inter-frame association stage also be error-prone, where an uncertainty problem …
View article: SLV: Spatial Likelihood Voting for Weakly Supervised Object Detection
SLV: Spatial Likelihood Voting for Weakly Supervised Object Detection Open
Based on the framework of multiple instance learning (MIL), tremendous works have promoted the advances of weakly supervised object detection (WSOD). However, most MIL-based methods tend to localize instances to their discriminative parts …
View article: HoMM: Higher-Order Moment Matching for Unsupervised Domain Adaptation
HoMM: Higher-Order Moment Matching for Unsupervised Domain Adaptation Open
Minimizing the discrepancy of feature distributions between different domains is one of the most promising directions in unsupervised domain adaptation. From the perspective of moment matching, most existing discrepancy-based methods are d…
View article: HoMM: Higher-order Moment Matching for Unsupervised Domain Adaptation
HoMM: Higher-order Moment Matching for Unsupervised Domain Adaptation Open
Minimizing the discrepancy of feature distributions between different domains is one of the most promising directions in unsupervised domain adaptation. From the perspective of distribution matching, most existing discrepancy-based methods…
View article: City brain: practice of large‐scale artificial intelligence in the real world
City brain: practice of large‐scale artificial intelligence in the real world Open
A city is an aggregate of a huge amount of heterogeneous data. However, extracting meaningful values from that data remains a challenge. City Brain is an end‐to‐end system whose goal is to glean irreplaceable values from big city data, spe…
View article: Towards Self-similarity Consistency and Feature Discrimination for Unsupervised Domain Adaptation
Towards Self-similarity Consistency and Feature Discrimination for Unsupervised Domain Adaptation Open
Recent advances in unsupervised domain adaptation mainly focus on learning shared representations by global distribution alignment without considering class information across domains. The neglect of class information, however, may lead to…
View article: See Extensively While Focusing on the Core Area for Pedestrian Detection
See Extensively While Focusing on the Core Area for Pedestrian Detection Open
Pedestrian detection attracts much attention from the academic community since it is an essential and significant component of autonomous driving. Despite many similarities with general object detection, pedestrian detection still has uniq…
View article: Previewer for Multi-Scale Object Detector
Previewer for Multi-Scale Object Detector Open
Most multi-scale detectors face a challenge of small-size false positives due to the inadequacy of low-level features, which have small receptive field sizes and weak semantic capabilities. This paper demonstrates independent predictions f…