A5017875019- Visual Attention and Saliency Detection
- Advanced Image and Video Retrieval Techniques
- Multimodal Machine Learning Applications
- Image and Video Quality Assessment
- Aesthetic Perception and Analysis
- Image Enhancement Techniques
- Face recognition and analysis
- Biometric Identification and Security
- Facial Nerve Paralysis Treatment and Research
- Electrochemical Analysis and Applications
- Electrochemical sensors and biosensors
- Analytical Chemistry and Sensors
- Cybercrime and Law Enforcement Studies
- Video Surveillance and Tracking Methods
- Advanced Algorithms and Applications
- Geoscience and Mining Technology
- Multi-Criteria Decision Making
- Generative Adversarial Networks and Image Synthesis
- Face and Expression Recognition
- Geomechanics and Mining Engineering
- Advanced Image Fusion Techniques
- Hate Speech and Cyberbullying Detection
- Mining Techniques and Economics
- Advanced Vision and Imaging
- Rough Sets and Fuzzy Logic
University of Science and Technology of China
2022-2024
Beijing Electronic Science and Technology Institute
2018-2019
In Suichang gold mine, the altered rock type deposits were cut by faults and joint fissures, leading to complex resource endowment characteristics, large changes in occurrence, a serious of ore vein branches great difficulty mining. order select suitable mining method for such difficult complicated orebody, multi-factor multi-index comprehensive evaluation system involving benefits, costs, safety other aspects was constructed using Pythagorean fuzzy sets TOPSIS method. Taking mine as an...
Image aesthetic quality assessment has been a relatively hot topic during the last decade. Most recently, comments type (aesthetic captions) proposed to describe general impression of an image using text. In this paper, we propose Aesthetic Attributes Assessment Images, which means attributes captioning. This is new formula assessment, predicts captions together with score each attribute. We introduce dataset named DPC-Captions contains up 5 one through knowledge transfer from full-annotated...
Face swapping has become a prominent research area in computer vision and image processing due to rapid technological advancements. The metric of measuring the quality most face methods relies on several distances between manipulated images source image, or target i.e., there are suitable known reference images. Therefore, is still gap accurately assessing interchange reference-free scenarios. In this study, we present novel no-reference assessment (NR-IQA) method specifically designed for...
<title>Abstract</title> For many years, the lead-zinc mine of Suichang Gold Mine has been mined by segmented open-field method, and mining above 500 m basically completed, forming a high-stage large-span columnar goaf group, which is very easy to produce roof fall, slope fragmentation instability collapse, then leading large-scale ground pressure disasters, causing surface subsidence endangering safety operations deep untouched ore bodies. In order better discuss stability goaf, Flac3D used...
Face-swapping techniques have advanced rapidly with the evolution of deep learning, leading to widespread use and growing concerns about potential misuse, especially in cases fraud. While many efforts focused on detecting swapped face images or videos, these methods are insufficient for tracing malicious users behind fraudulent activities. Intrusive watermark-based approaches also fail trace unmarked identities, limiting their practical utility. To address challenges, we introduce...
It is difficult and challenging to evaluate the aesthetics quality of images from multiple angles. Since humans' perception comes many factors, integrated image aesthetic assessment cannot be easily summarized by few attributes. A comprehensive evaluation supposed predict attributes across not only one dataset. This requires model have high accuracy, but also strong generalization ability, resulting in a better prediction on models datasets. Recent work shows that deep convolution neural...
The task of aesthetic quality assessment is complicated due to its subjectivity. In recent years, the target representation image has changed from a one-dimensional binary classification label or numerical score multi-dimensional distribution. According current methods, ground truth distributions are straightforwardly regressed. However, subjectivity aesthetics not taken into account, that say, psychological processes human beings consideration, which limits performance task. this paper, we...
Aesthetic assessment of images can be categorized into two main forms: numerical and language assessment. In this paper, we propose a new task aesthetic assessment: visual question answering (AVQA) images. We use from www.flickr.com. The objective QA pairs are generated by the proposed attributes analysis algorithms. Moreover, introduce subjective that converted labels sentiment large-scale pre-train models. build first dataset, AesVQA, contains 72,168 high-quality 324,756 questions. This is...
This paper presents the summary report on DeepFake Game Competition Visual Realism Assessment (DFGC-VRA). Deep-learning based face-swap videos, also known as deepfakes, are becoming more and realistic deceiving. The malicious usage of these videos has caused wide concerns. There is a ongoing deepfake game between its creators detectors, with human in loop. research community been focusing automatic detection fake but assessment their visual realism, perceived by eyes, still an unexplored...
An album face coding system is developed by python based on the human joint Bayesian algorithm that improved and applied in recognition. Experiments show can work real time with high accuracy.
Image aesthetic quality assessment has been a relatively hot topic during the last decade. Most recently, comments type (aesthetic captions) proposed to describe general impression of an image using text. In this paper, we propose Aesthetic Attributes Assessment Images, which means attributes captioning. This is new formula assessment, predicts captions together with score each attribute. We introduce dataset named \emph{DPC-Captions} contains up 5 one through knowledge transfer from...
Image aesthetic quality assessment is popular during the last decade. Besides numerical assessment, nature language (aesthetic captioning) has been proposed to describe generally impression of an image. In this paper, we propose attribute which attributes captioning, i.e., assess such as composition, lighting usage and color arrangement. It a non-trivial task label comments attributes, limit scale corresponding datasets. We construct novel dataset, named DPC-CaptionsV2, by semi-automatic...
Aesthetic assessment of images can be categorized into two main forms: numerical and language assessment. Aesthetics caption photographs is the only task aesthetic that has been addressed. In this paper, we propose a new assessment: visual question answering (AVQA) images. If give aesthetics, model predict answer. We use from \textit{www.flickr.com}. The objective QA pairs are generated by proposed attributes analysis algorithms. Moreover, introduce subjective converted labels sentiment...