S
S.M.R. Soroushmehr
Researcher at University of Michigan
Publications - 29
Citations - 902
S.M.R. Soroushmehr is an academic researcher from University of Michigan. The author has contributed to research in topics: Segmentation & Convolutional neural network. The author has an hindex of 13, co-authored 29 publications receiving 660 citations.
Papers
More filters
Book ChapterDOI
A Game Theoretic Predictive Modeling Approach to Reduction of False Alarm
Fatemeh Afghah,Abolfazl Razi,S.M.R. Soroushmehr,Somayeh Molaei,Hamid Ghanbari,Kayvan Najarian +5 more
TL;DR: A model based on coalition game that considers the inter---features mutual information which results in gaining the accuracy of the classification is proposed which shows the superior performance of the proposed method compared to other existing methods.
Proceedings ArticleDOI
Hardware image assessment for wireless endoscopy capsules
Mohammad Amin Khorsandi,Nader Karimi,Shadrokh Samavi,Mohsen Hajabdollahi,S.M.R. Soroushmehr,Kevin R. Ward,Kayvan Najarian +6 more
TL;DR: An architecture that could be used for the assessment of endoscopy images that has low complexity and is appropriate for a real-time application is proposed.
Posted Content
Low complexity convolutional neural network for vessel segmentation in portable retinal diagnostic devices
Mohsen Hajabdollahi,Reza Esfandiarpoor,S.M.R. Soroushmehr,Nader Karimi,Shadrokh Samavi,K Najarian +5 more
TL;DR: A simplification approach is proposed for CNNs based on combination of quantization and pruning and shows that the simplified network is able to segment retinal vessels with acceptable accuracy and low complexity.
Proceedings ArticleDOI
Radon transform inspired method for hand gesture recognition
Mohammad Amin Khorsandi,Nader Karimi,S.M.R. Soroushmehr,Mohsen Hajabdollahi,Shadrokh Samavi,Kevin R. Ward,Kayvan Najarian +6 more
TL;DR: A new hand gesture recognition method that is Simplicity and robustness against rotation, scaling and position and also having no complex mathematical calculation are advantages of this work.
Proceedings ArticleDOI
Robust catheter identification and tracking in X-ray angiographic sequences
H. R. Fazlali,Nader Karimi,S.M.R. Soroushmehr,Shadrokh Samavi,Brahmajee K. Nallamothu,Harm Derksen,Kayvan Najarian +6 more
TL;DR: This paper proposes a fully automatic method for catheter detection and tracking during the whole angiography sequence with a vesselness map, and is tested on 25 X-ray angiographs where a precision of 0.9597 is achieved.