Posts by Collection

projects

Robotic Arm Kinematics

A ROS-based project implementing forward and inverse kinematics for a robotic arm

Turtlebot Path Planning

Implemented Bug1 and Bug2 algorithms for the Turtlebot3 Burger robot

Image Panorama Stitching

Panorama creation from multiple images using Python, OpenCV, and SIFT algorithm

Smart Class IoT

A project enhancing classroom security with human detection, door access control, and attendance tracking

Fake Image Detection

This project used CNNs and image processing to detect fake face images, achieving a 0.94 f1-score

FAQ Answering Bot (Bachelor’s Thesis Project)

A lightweight chatbot to answer frequently asked questions using Siamese Networks

Latent Space Manipulation

Edited face features using GANs, tackling entanglement, artifacts, and identity-preserving encoding

Virtual Try-On

Used custom GANs to swap garments on a person, solving issues with resolution and limited data, trained distributedly

Physics-Aware LiDAR Simulation

This project develops a physics-based LiDAR simulation using deep learning to improve accuracy.

publications

Vim4Path: Self-Supervised Vision Mamba for Histopathology Images

Published in IEEE/CFV Conference on Computer Vision and Pattern Recognition - CVMI Workshop, 2024

This paper introduces the Vision Mamba (Vim) architecture, outperforming Vision Transformers (ViT) in computational pathology on the Camelyon16 dataset.

Recommended citation: A. Nasiri-Sarvi, V. Q.-H. Trinh, H. Rivaz, and M. S. Hosseini, “Vim4path: Self-supervised vision mamba for histopathology images,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops
Download Paper

Vision mamba for classification of breast ultrasound images

Accepted in International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), Deep-Brea3th Workshop, 2024

This paper shows that Mamba-based models outperform CNNs and ViTs on breast ultrasound datasets, achieving a 1.98% AUC and 5.0% accuracy improvement, while effectively capturing long-range dependencies.

Recommended citation: A. Nasiri-Sarvi, M. S. Hosseini, and H. Rivaz, “Vision mamba for classification of breast ultrasound images,” arXiv preprint arXiv:2407.03552, 2024.
Download Paper

talks

Vim4Path: Self-Supervised State Space Modeling for Histopathology Images

Published: June 18, 2024

https://cvmi-workshop.github.io/program.html

Ali Nasiri-Sarvi