A5029719997- Adversarial Robustness in Machine Learning
- Anomaly Detection Techniques and Applications
- Digital Media Forensic Detection
- Spectroscopy and Quantum Chemical Studies
- Domain Adaptation and Few-Shot Learning
- Music and Audio Processing
- Human Pose and Action Recognition
- Biopolymer Synthesis and Applications
- Speech and Audio Processing
- Advanced Neural Network Applications
- Free Radicals and Antioxidants
- Chemical Synthesis and Reactions
- Video Analysis and Summarization
- Computational Drug Discovery Methods
- High Entropy Alloys Studies
- High-Temperature Coating Behaviors
- Video Surveillance and Tracking Methods
- Catalysis and Hydrodesulfurization Studies
- Environmental Toxicology and Ecotoxicology
- Multimodal Machine Learning Applications
- Air Quality and Health Impacts
- Toxic Organic Pollutants Impact
- Additive Manufacturing Materials and Processes
Chinese Academy of Sciences
2020-2024
Institute of Information Engineering
2022-2024
University of Chinese Academy of Sciences
2013-2024
Institute of Tibetan Plateau Research
2020
Xiangtan University
2006
Video-text retrieval is an emerging stream in both computer vision and natural language processing communities, which aims to find relevant videos given text queries. In this paper, we study the notoriously challenging task, i.e., Unsupervised Domain Adaptation Retrieval (UDAVR), wherein training testing data come from different distributions. Previous works merely alleviate domain shift, however overlook pairwise misalignment issue target domain, there exist no semantic relationships...
Person Re-IDentification (ReID) aims at retrieving images of the same person across multiple camera views. Despite its popularity in surveillance and public safety, leakage identity information is still risk. For example, once obtaining illegal access to ReID systems, malicious user can accurately retrieve target person, leading exposure private information. Recently, some pioneering works protect with adversarial examples by adding imperceptible perturbations images. However, this paper, we...
Video is a natural source of multi-modal data with intrinsic correlations between different modalities, such as objects, motions and captions. Though intuitive, inherent supervision has not been well explored in previous video-audio retrieval works. Besides, existing methods exploit the video stream audio seperately, whereas ignoring mutual interactions them. In this paper, we propose two-stream model named Multi-modal Aggregation Co-attention network (MAC), which processes inputs...
Although image-dependent adversarial attacks have been studied, the more challenging image-agnostic attack for deep hashing remains an unexplored territory. In this paper, we take first attempt on efficient and malicious targeted universal (TUAA) hashing. When previous are directly applied to TUAA task, they usually face two main issues. Firstly, existing anchor code generation methods generate with inferior representative semantic-preserving ability. Secondly, simply minimize distance...
γ-polyglutamic acid (PGA) was firstly used for recovery and immobilization of Pseudomonas delafieldii R-8 cells in biodesulfurization process. The PGA mediated flocculation combined with celite adsorption to immobilize the as-prepared immobilized showed a high initial specific desulfurization rate at 0.243 mmol l -1 h good stability 95% activity remained after 6 batches processes.
With the extensive utilization of deep hashing, there exists a surging interest in studying adversarial attacks against it. Previous methods have demonstrated superior white-box attack performance hashing. However, more challenging and realistic targeted black-box has not yet been explored sufficiently, which will result an over-estimation on model robustness. In this paper, we focus based transferability, propose novel Targeted Transferable Attack method hashing with Generative Adversarial...