Research
We are committed to developing novel and efficient deep and machine learning models that enhance performance, interpretability and trustworthiness, and improve impact and accessibility. Our solutions are designed to contribute to the academic advancements, strategic growth and platform investments of Canadian industries. AiX Lab is dedicated to cultivating highly skilled scholars with cutting-edge AI and equipping them to contribute to various industry sectors.
Projects
Automotive Research
RL for Autonomous Navigation in Self-driving Cars
Deep Reinforcement Learning (DRL) plays a fundamental role in autonomous driving applications. This research introduces sample-efficient DRL models for decision-making, path planning and automatic navigation of AVs in complex scenarios.
Autonomous Driving System for UGV
This project concerns the development and deployment of an autonomous driving system for an unmanned ground vehicle, called Jackal. The Jackal vehicle will be used in smart city applications, such structural evaluation. A LiDAR senor was used to provide real-world data for driving system.
Scene Perception Models in Self Driving Cars
This project aims to enhance real-time multi-object detection capabilities for self-driving cars, specifically operating effectively in various adverse weather conditions (e.g., rain, fog, and snow), which can significantly impair the accuracy and reliability of object detection systems in autonomous vehicles.
Driver Attention
This research focuses on driver's visual and manual distraction. Capturing driver’s video through camera, attention level is analyzed. Key objective is to design reliable and compact models to facilitate gaze estimation and action recognition in DMS.
Driver Impairment
This research focuses on addressing driver fatigue and the influence of alcohol and drugs on driving behavior. Through innovative solutions and advanced technologies, we aim to uncover key biomarkers and behavioral indicators for real-time impairment monitoring and contributing to a safer driving.
Driver State
This project is dedicated to understanding the dynamic assessment of driver's emotion, well-being and cognitive workload and enhance the overall driving experience. The objective is to shape the future of transportation with a focus on holistic driver well-being and mental resilience.
Adaptive Handover System
An adaptive handover system is a novel idea aimed at optimizing the control transfer between driver and automated system. This requires dynamically adjusting a TOR based on a real-time analysis of both driver's state and road/traffic conditions to perform the most appropriate action. This could mean altering a request's modality, timing, or location. Such an adaptive approach promises to enhance safety by mitigating the risks associated with a handover system.
Clinical Research
AI Screener fro Children with Listening Difficulties
The main objective of this project is to develop a novel smart screener framework for screening children with listening difficulties by analyzing the ABR signals using AI technologies. The goal is to analyze and identify a stopping criterion to obtain a robust ABR, learn patterns from data and create a non-invasive smart screening tool to help clinicians in diagnosis of hearing difficulties.
AI-Based Musical Intervention to Improve Emotion
Mental disorders have a significant influence on the daily activities of Canadians. Musical intervention can provide a non-invasive treatment through changing emotional state and creating positive mood. The main objective of this project is a long-term solution for musical intervention through an optimized machine learning framework for a real-time emotion recognition and musical intervention system integrated in an empathetic speaker. During music play, the emotional influence will be detected from EEG and the music database will be customized.
Agriculture Research
Vision based growth monitoring
Leveraging AI and computer vision technologies, we are dedicated to transforming the assessment of the growth dynamics of crops and plants. Our vision-based approach enables precise and real-time monitoring of key indicators, contributing to enhanced crop management practices. From tracking plant development to identifying potential stress factors, our research aims to empower farmers with actionable insights for optimized yields and sustainable farming.
Selected Publications
Book Chapter
S. Nikan, M. Ahmadi, “Recognition of human faces under different degradation conditions,” in Trends in Digital Signal Processing, 1st ed. Singapore, Singapore: Pan Stanford Publishing Pte. Ltd., vol. 1, ch. 11, pp. 333-356, 2015.
Journal
R. Yahyaabadi and S. Nikan, "An Explainable Attention Zone Estimation for Level 3 Autonomous Driving," IEEE Access, vol. 11, pp. 93098 - 93110, 2023.
Y. Ma, V. Sanchez, S. Nikan, D. Upadhyay, B. Atote, T. Guha, "Real-Time Driver Monitoring Systems through Modality and View Analysis," arXiv preprint arXiv:2210.09441, 2022.
S. Nikan, K. Van Osch, M. Bartling, D. G. Allen, S. A. Rohani, B. Connors, S. K. Agrawal, H. M. Ladak, PWD-3DNet: A deep learning-based fully automated segmentation of multiple structures on temporal bone CT scans, IEEE Transactions on Image Processing, 2020, vol. 30, pp. 739-753.
R. Yahyaabadi and S. Nikan, " Efficient 2D/3D gaze estimation using TGGNet: A transformer graph approach," IEEE Transactions on Image Processing, under review, 2023.
H. Hassani, S. Nikan, A. Shami, "Improved exploration-exploitation trade-off through adaptive prioritized experience replay," IEEE Transactions on Artificial Intelligence, under review, 2023.
M. Nabipour and S. Nikan, Action unit analysis for monitoring drivers’ emotional states, IEEE Sensors Journal, under review, 2023.
H. Hassani, S. Nikan, A. Shami, "Traffic navigation via reinforcement learning with Episodic-Guided prioritized experience replay," IEEE Transactions on Intelligent Vehicles, under review, 2023.
S. Nikan, K. Van Osch, M. Bartling, D. G. Allen, S. A. Rohani, B. Connors, S. K. Agrawal, H. M. Ladak, PWD-3DNet: A deep learning-based fully automated segmentation of multiple structures on temporal bone CT scans, IEEE Transactions on Image Processing, 2020, vol. 30, pp. 739-753.
S. Nikan, M. Ahmadi, A modified technique for face recognition under degraded conditions, Journal of Visual Communication and Image Representation, 2018, 55, pp. 742-755.
F. Gwadry-Sridhar, S. Nikan, A. Hamou, SJ. Seung, T. Petrella, AM. Joshua, S. Ernst, N. Mittmann, Resource utilization and costs of managing patients with advanced melanoma: a Canadian population-based study, Current Oncology, 2017, vol. 24, no. 3, pp. 168–175.
S. Nikan, M. Ahmadi, Recognition of human faces in the presence of incomplete information, International Journal on Advances in Software, 2015, vol. 8, no. 3&4, pp. 450-456.
S. Nikan, M. Ahmadi, Performance evaluation of different feature extractors and classifiers for recognition of human faces with low resolution images, International Journal of Intelligent Systemsand Applications in Engineering, 2014, vol. 3, no. 2, pp. 72-77.
S. Nikan, M. Ahmadi, Local gradient-based illumination invariant face recognition using LPQ and multi-resolution LBP fusion, IET Image Processing, 2015, vol. 9, no. 1, pp. 12-21.
Conference
R. Yahyaabadi and S. Nikan, "Skeleton-based driver action recognition using ResGGCNN," in ISSCS, Iasi, Romania, pp. 1-4, 2023.
S. Nikan and D. Upadhyay, "Appearance-Based Gaze Estimation for Driver Monitoring," in NeurIPS, pp. 127-139, 2023.
Y. Ma, V. Sanchez, S. Nikan, D. Upadhyay, B. Atote, T. Guha, "Robust Multiview Multimodal Driver Monitoring System Using Masked Multi-Head Self-Attention," in CVPR, pp. 2616-2624, 2023.
M. Nabipour and S. Nikan, "A Deep Learning-Based Remote Plethysmography with the Application in Monitoring Drivers’ Wellness," in ISSCS, Iasi, Romania, pp. 1-4, 2023.
M. Mohseni, S. Nikan, A. Shami. AI-based Traffic Forecasting in 5G network, in CCECE2022, Accepted, September 2022.
S. Nikan, S. K. Agrawal, H. M. Ladak, Fully Automated Segmentation of the Temporal Bone from Micro-CT using Deep Learning, in Proc Biomedical Applications in Molecular, Structural, and Functional Imaging (SPIE’20), Houston, TX, United States, Feb 2020.
M. Samadi, H. Kharrati, M.A. Badamchizadeh, H. Hassani*, S. Nikan. (2023). Using liquid time-constant network for diagnosing faults and cyber-attacks in smart grids. International Conference on Engineering and Emerging Technologies (ICEET 2023), in press.
S. Nikan, S. K. Agrawal, H. M. Ladak, Fully Automated Segmentation of the Temporal Bone from Micro-CT using Deep Learning, in SPIE’20, Houston, TX, United States, Feb 2020.
S. Nikan, S. Agrawal, H. Ladak, Automated multi-structure deep segmentation of micro-CT images of temporal bone, LHRD’19, 2019.
S. Nikan, F. Gwadry-Sridhar, M. Bauer, Pattern recognition application in ECG arrhythmia classification, in HEALTINF’17, Porto, Portugal, Feb 2017, pp. 48-56.
S. Nikan, F. Gwadry-Sridhar, M. Bauer, Machine learning application to predict the risk of coronary artery atherosclerosis, in CSCI’16, Las Vegas, USA, Dec 2016, pp. 34-39.
S. Nikan, M. Ahmadi, Partial Face Recognition Based on Template Matching, in SITIS’15, Bangkok, Thailand, Nov 2015, pp. 160-163.
S. Nikan, M. Ahmadi, “Partial face recognition based on template matching,” in SITIS’15, Bangkok, Thailand, pp. 160-163, Nov 2015.