A5000728600- Advanced Image and Video Retrieval Techniques
- Video Analysis and Summarization
- Topic Modeling
- Music and Audio Processing
- Human Pose and Action Recognition
- Machine Learning and Data Classification
- Image Retrieval and Classification Techniques
- Botanical Research and Applications
- Robot Manipulation and Learning
- Text and Document Classification Technologies
- Sports Analytics and Performance
- Advanced Graph Neural Networks
- Fault Detection and Control Systems
- Algorithms and Data Compression
- Data Stream Mining Techniques
- Botany, Ecology, and Taxonomy Studies
- Expert finding and Q&A systems
- Smart Grid Energy Management
- Domain Adaptation and Few-Shot Learning
- Recommender Systems and Techniques
- Multimodal Machine Learning Applications
- Plant Diversity and Evolution
Sharif University of Technology
2012-2023
Semantic video analysis and automatic concept extraction play an important role in several applications; including content-based search engines, indexing, summarization. As the Bayesian network is a powerful tool for learning complex patterns, novel network-based method proposed event detection summarization soccer videos. The includes efficient algorithms shot boundary detection, view classification, mid-level visual feature extraction, construction of related network. contains three main...
Multi-label classification is a learning task in which each data sample can belong to more than one class. Until now, some methods that are based on reducing the dimensionality of label space have been proposed. However, these not used specific properties for this purpose. In paper, we intend find hidden both input feature vectors and embedded. We propose modified Non-Negative Matrix Factorization (NMF) method suitable decomposing matrix finding proper by feature-aware approach. consider...
Transformers have become widely used in modern models for various tasks such as natural language processing and machine vision. This paper proposes Gransformer, an algorithm generating graphs based on the Transformer. We extend a simple autoregressive Transformer encoder to exploit structural information of given graph through efficient modifications. The attention mechanism is modified consider presence or absence edges between each pair nodes. also introduce graph-based familiarity measure...
Voluminous amount of videos in today's world has made the subject objective (or semi-objective) classification to be very popular. Among various descriptors used for video classification, SIFT and LIFT can lead highly accurate classifiers. But, descriptor does not consider motion is time-consuming. In this paper, a robust semi-supervised based on content proposed. It holds benefits overcomes their shortcomings some extent. For extracting descriptor, first extracted keypoints are then...
In large databases, availability of labeled training data is mostly prohibitive in classification. Semi-supervised algorithms are employed to tackle the lack problem. Video databases epitome for such a scenario; that why semi-supervised learning has found its niche it. Graph-based methods promising platform video Based on multiview characteristic data, different features have been proposed (such as SIFT, STIP and MFCC) which can be utilized build graph. this paper, we new classification...