A5012933659- Advanced Vision and Imaging
- Video Surveillance and Tracking Methods
- Advanced Image and Video Retrieval Techniques
- Video Analysis and Summarization
- Human Pose and Action Recognition
- Image and Signal Denoising Methods
- Video Coding and Compression Technologies
- Face and Expression Recognition
- Robotics and Sensor-Based Localization
- Image Retrieval and Classification Techniques
- Remote-Sensing Image Classification
- Advanced Data Compression Techniques
- Advanced Neural Network Applications
- Medical Image Segmentation Techniques
- Infrared Target Detection Methodologies
- Anomaly Detection Techniques and Applications
- Advanced Image Processing Techniques
- Adversarial Robustness in Machine Learning
- Image Enhancement Techniques
- Robotic Path Planning Algorithms
- Advanced Image Fusion Techniques
- Visual Attention and Saliency Detection
- Face recognition and analysis
- Hand Gesture Recognition Systems
- Image Processing Techniques and Applications
Sharif University of Technology
2016-2025
Yazd University
2019-2020
Institute for Research in Fundamental Sciences
2018
University of Aveiro
2018
Islamic Azad University, Isfahan
2009-2012
Islamic Azad University, Tehran
2010
Queensland University of Technology
2002-2003
Visual target tracking is one of the most sought-after yet challenging research topics in computer vision. Given ill-posed nature problem and its popularity a broad range real-world scenarios, number large-scale benchmark datasets have been established, on which considerable methods developed demonstrated with significant progress recent years -- predominantly by deep learning (DL)-based methods. This survey aims to systematically investigate current DL-based visual methods, datasets,...
The interest in action and gesture recognition has grown considerably the last years. In this paper, we present a survey on current deep learning methodologies for image sequences. We introduce taxonomy that summarizes important aspects of approaching both tasks. review details proposed architectures, fusion strategies, main datasets, competitions. summarize discuss works so far with particular how they treat temporal dimension data, discussing their features identify opportunities...
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...
As the panoramic x-ray is most common extraoral radiography in dentistry, segmentation of its anatomical structures facilitates diagnosis and registration dental records. This study presents a fast accurate method for automatic mandible x-rays. In proposed four-step algorithm, superior border extracted through horizontal integral projections. A modified Canny edge detector accompanied by morphological operators extracts inferior body. The exterior borders ramuses are contour tracing based on...
This paper presents a feature-based license plate localization algorithm that copes with multi-object problem in different image capturing conditions. The proposed is robust against illumination, shadow, scale, rotation, and weather condition. It extracts candidates using edge statistics morphological operations removes the incorrect according to determined features of plates. We have formed rather complete database 269 images successfully detecteds accurate location plates 96.5% cases,...
Hand gesture recognition (HGR) from sequences of depth maps is a challenging computer vision task because the low inter-class and high intra-class variability, different execution rates each gesture, articulated nature human hand. In this paper, multilevel temporal sampling (MTS) method first proposed that based on motion energy keyframes sequences. As result, long, middle, short are generated contain relevant information. The MTS results in increasing similarity while raising...
matching. In this system main stage is the isolation of license plate, from digital image car obtained by a camera under different circumstances such as illumination, slop, distance, and angle. The algorithm starts with preprocessing signal conditioning. Next plate localized using morphological operators. Then template matching scheme will be used to recognize digits characters within plate. tested on Iranian images, performance was 97.3% correct plates identification localization 92%...
There is growing need for robots that can interact with people in everyday situations. For service robots, it not reasonable to assume one pre-program all object categories. Instead, apart from learning a batch of labelled training data, should continuously update and learn new categories while working the environment. This paper proposes cognitive architecture designed create concurrent 3D category recognition an interactive open-ended manner. In particular, this provides automatic...
The 3D point clouds are increasingly being used in various application including safety-critical fields. It has recently been demonstrated that deep neural networks can successfully process clouds. However, these be misclassified via adversarial attacks intentionality designed to perturb some cloud's features. These misclassifications may due the network's overreliance on features with unnecessary information training sets. As such, identifying by classifiers and removing from data improve...
Pomegranate is a temperature-sensitive fruit during postharvest storage. If exposed to cold temperatures above its freezing point for long time, it will suffer from stress. Failure pay attention the symptoms that may occur storage result in significant damage. Identifying pomegranates susceptible damage timely manner requires considerable skill, time and cost. Therefore, non-destructive real-time methods offer great benefits commercial producers. To this end, purpose of study identification...
Despite significant progress in deep learning-based optical flow methods, accurately estimating large displacements and repetitive patterns remains a challenge. The limitations of local features similarity search used these algorithms contribute to this issue. Additionally, some existing methods suffer from slow runtime excessive graphic memory consumption. To address problems, paper proposes novel approach based on the RAFT framework. proposed Attention-based Feature Localization (AFL)...
Fingerprints have been used as unique identifiers of individuals for a very long time. As fingerprint databases are characterized by their large size and may contain noisy distorted images, an efficient representation the images is essential reliable identification. Considering fingerprints sample from non-stationary processes with flow patterns, we propose here robust technique to extract features. The properties textures enhance improve fidelity ridges extracted enhanced foreground areas...