A5100640861- Music and Audio Processing
- Music Technology and Sound Studies
- Diverse Musicological Studies
- Speech and Audio Processing
- Advanced Neural Network Applications
- Advanced Clustering Algorithms Research
- Complex Network Analysis Techniques
- Video Surveillance and Tracking Methods
- Cryptographic Implementations and Security
- Privacy-Preserving Technologies in Data
- Privacy, Security, and Data Protection
- Graph theory and applications
- Physical Unclonable Functions (PUFs) and Hardware Security
- Machine Learning and Data Classification
- Text and Document Classification Technologies
- Limits and Structures in Graph Theory
- Energy Efficient Wireless Sensor Networks
- Advanced Graph Theory Research
- Traffic Prediction and Management Techniques
- E-commerce and Technology Innovations
- Advanced Image and Video Retrieval Techniques
- Autonomous Vehicle Technology and Safety
- Face and Expression Recognition
- Advanced Sensor and Control Systems
- Advanced Chemical Sensor Technologies
Shandong Jiaotong University
2019-2024
Queen Mary University of London
2020-2021
China University of Petroleum, East China
2021
Jiangnan University
2017
Northeastern University
2017
The ongoing energy crisis has underscored the urgent need for energy-efficient materials with high utilization efficiency, prompting a surge in research into organic compounds due to their environmental compatibility, cost-effective processing, and versatile modifiability. To address experimental costs time-consuming nature of traditional trial-and-error methods discovery highly functional compounds, we apply 3D transformer-based molecular representation learning algorithm construct...
Watermarking as a novel intellectual property (IP) protection technique can protect field-programmable gate array IPs from infringement. However, existing watermarking techniques may give away sensitive information during the public verification, which enables malicious verifiers or third parties to remove embedded watermark and resell design. Current zero-knowledge verification schemes address leakage issue but are vulnerable embedding attacks, makes them ineffective in preventing...
Automatic Music Transcription (AMT) is usually evaluated using low-level criteria, typically by counting the number of errors, with equal weighting. Yet, some errors (e.g. out-of-key notes) are more salient than others. In this study, we design an online listening test to gather judgements about AMT quality. These take form pairwise comparisons transcriptions same music pairs different systems. We investigate how these correlate benchmark metrics, and find that although they match in many...
Privacy protection problem is one of the most concerning issues related to Location-Based Services (LBS) in our daily life. LBS often requires anonymizing customer's trajectory data. Currently available methods for anonymity assume an entire as anonymous unit, which may lead low efficiency due massive amount data, especially customers travel through a long road. Considering people's routine activities, starting and ending locations trip uncover user's request intent, exposure user privacy....
Research on automatic music transcription has largely focused multi-pitch detection; there is limited discussion how to obtain a machine- or human-readable score transcription. In this paper, we propose method for joint detection and polyphonic piano music. The outputs of our system include both piano-roll representation (a descriptive transcription) symbolic musical notation prescriptive transcription). Unlike traditional methods that further convert MIDI transcriptions into scores, use...
<title>Abstract</title> Vehicle detection algorithms are integral to intelligent traffic management and AI-assisted driving systems. However, the complexity variability of scenarios present significant challenges. Spatial pyramid pooling enhances a model’s ability capture multi-scale contextual information by offering different levels feature representation, which is particularly beneficial in complex scenarios. In this study, we extend spatial Transformer-based models introduce linear...
<title>Abstract</title> Model compression reduces the size of large neural networks while preserving performance. Progressive Blockwise Knowledge Distillation (PBKD) is a model technique that achieves lightweight and efficient student by progressively replacing teacher model's subnet-work block sequence with subnetwork through block-by-block opti-mization process. We propose novel approach to designing blocks emphasizes greater deeper within network in PBKD framework. Compared previous...
Vehicle detection algorithms are essential for intelligent traffic management and autonomous driving systems. Current vehicle largely rely on deep learning techniques, enabling the automatic extraction of image features through convolutional neural networks (CNNs). However, in real scenarios, relying only a single feature unit makes it difficult to fully understand information scenario, thus affecting effect. To address this issue, we propose lightweight algorithm based Mamba_ViT. First,...
This technical report gives a detailed, formal description of the features introduced in paper: Adrien Ycart, Lele Liu, Emmanouil Benetos and Marcus T. Pearce. "Investigating Perceptual Validity Evaluation Metrics for Automatic Piano Music Transcription", Transactions International Society Information Retrieval (TISMIR), Accepted, 2020.
Affinity propagation algorithm is a new powerful and effective clustering method. One of the major problems in determination optimal number clusters. In this paper, particle swarm optimization utilized to cope with problem by using parameter p as each Silhouette index fitness, which can search for value determine clusters automatically. Moreover, information overlap main drawback affinity dealing complex structure or high dimensional data clustering. Hence enhanced Locality preserving...
Artificial Intelligence (AI) technologies such as deep learning are evolving very quickly bringing many changes to our everyday lives. To explore the future impact and potential of AI in field music sound a doctoral day was held between Queen Mary University London (QMUL, UK) Sciences et Technologies de la Musique du Son (STMS, France). Prompt questions about current trends were generated by academics from QMUL STMS. Students two institutions then debated these questions. This report...
User behavior recognition using sensory data has become an active field of research in the domain pervasive and mobile computing. The Principal Component Analysis (PCA) is a common method for feature selection. To obtain best description classification characteristics different behaviors, algorithm based on Regularized Mutual Information (RMIPCA) presented. new introduces category information, uses sum regularized mutual information matrices between features under to replace covariance...
Let $G$ be a connected graph, the principal ratio of is maximum and minimum entries its Perron eigenvector. In 2007, Cioab\v Gregory conjectured that among all graphs on $n$ vertices, kite graph attains ratio. 2018, Tait Tobin confirmed conjecture for sufficientlty large $n$. this article, we show true $n\geq 5000$.