Welcome


Welcome to the home page for University of Tehran Computational Audio-Vision Lab.

 

Our group develops new optimization and machine learning tools and use them in diverse applications primarily with emphasis on machine vision. Among optimization methods, we are particularly interested in manifold optimization.


Highlights

 

Raabin-WBC Data is available.

MixEst Toolbox website is up and running.
 

 

News (2023)

 

Our paper on FRMDN: flow-based recurrent mixture density network was published in Expert Systems with Applications.
 

Our paper on Stereo-RSSF: stereo robust sparse scene-flow estimation was published in The Visual Computer.
 

Our paper on The optimal triangulation method is not really optimal was published in IET Image Processing.
 

Mohammad Tavakili and Mahdi Mousavi successfully defended their Bachelor thesis.
 

Ali Izadi, Mohammad Mahdi Mehmanchi, Mohammad Ali Adelimanesh, Maedeh Zarvandi, Erfan Mirzaei, Banafsheh Karimian and Amirhossein Mesbah successfully defended their Master thesis.
 

Our paper on Online relational tracking with camera motion suppression was published in Journal of Visual Communication and Image Representation.
 

Our paper on Self-attention presents low-dimensional knowledge graph embeddings for link prediction was published in Knowledge-Based Systems.

Hadi Asheri and Mohammad Nasseri successfully defended their PhD thesis.


 

 

News (2022)

 

Reshad Hosseini was the computer vision instructor in AI and Python summer bootcamp of Hamrah Academy.
 

Our paper on Multiple-solutions RANSAC for finding axes of symmetry in fragments of objects was published in Pattern Recognition.
 

Our paper on A large dataset of white blood cells containing cell locations and types, along with segmented nuclei and cytoplasm was published in Scientific Reports.
 

Peyman Baghershahi, Afsaneh Hasabebrahimi, Bahareh Kaviani, Hamidreza Soltabalizadeh, Fatemeh Noorzad and Mohammad Pasandeh successfully defended their Master theses.
 

Elham Abolhasani and Narjer Noorzad successfully defended their Bachelor projects.
 

Mahsa Masoud and Omid Vaheb successfully defended their Bachelor Projects. They were awarded best student thesis Awards. Congratulation!
 

News (2021)

 

Our paper on Parsinorm: A persian toolkit for speech processing normalization got the best conference paper award.
 

Our paper on New segmentation and feature extraction algorithm for classification of white blood cells in peripheral smear images was published in Scientific Reports.
 

Our paper on A new EM algorithm for flexibly tied GMMs with large number of components was published in Pattern Recognition.
 

Our paper on Vector Transport Free Riemannian LBFGS for Optimization on Symmetric Positive Definite Matrix Manifolds was published in Asian Conference on Machine Learning.
 

Aida Khozaei and Seyed-Mehdi Nasiri defended their PhD theses.
 

Mohammad Hamed and Alireza Hariri successfully defended their Master theses.
 

Mohammad Hossein Hosseini, Tina Behzad and Amin Fadaeinejad successfully defended their Bachelor projects.
 

News (2020)

 

Our book chapter on Recent advances in stochastic Riemannian optimization was published in Handbook of Variational Methods for Nonlinear Geometric Data.
 

Our paper on Novel Parameterization for Gauss–Newton Methods in 3-D Pose Graph Optimization was published in IEEE Transactions on Robotics.
 

Our paper on Early screening of autism spectrum disorder using cry features was published in PloS One.
 

Sajjad Abdi, Fatemeh Pooyan, Amirhossein Kharatian, Ramin Yadegar and Milad Ghorbani successfully defended their Master theses.
 

News (2019)

 

Mahdi Darabi and Mohammad Ebrahimi successfully defended their Master theses.
 

Reshad Hosseini gave a talk in Amirkabir Artificial Intelligence Summer Summit.
 

Our paper on Exploiting Generalization in the Subspaces for Faster Model-Based Reinforcement Learning was published in IEEE Transactions on Neural Networks and Learning Systems.
 

Our paper on Clustering subspace generalization to obtain faster reinforcement learning was published in Evolving Systems.
 

Alireza Ganjdansh, Peiman Mohseni and Amirali Darbandsari successfully defended their Bachelor projects.
 

Our paper on Riemannian Optimization for GMM was accepted in Mathematical Programming.
 

Reshad Hosseini gave a talk in the tutorial session of CiDas Conference. Hadi Asheri and Pourya Habibzadeh organized Pytorch workshop in the conference.
 

Zahra Habibollahi successfully defended her Bachelor Project. She was awarded best student thesis Award. Congratulation!