M
Mert R. Sabuncu
Researcher at Cornell University
Publications - 148
Citations - 12746
Mert R. Sabuncu is an academic researcher from Cornell University. The author has contributed to research in topics: Computer science & Convolutional neural network. The author has an hindex of 24, co-authored 118 publications receiving 9297 citations. Previous affiliations of Mert R. Sabuncu include Massachusetts Institute of Technology & Harvard University.
Papers
More filters
Posted Content
NeuroGen: activation optimized image synthesis for discovery neuroscience.
Zijin Gu,Keith Jamison,Meenakshi Khosla,Emily J. Allen,Yihan Wu,Thomas Naselaris,Kendrick Kay,Mert R. Sabuncu,Amy Kuceyeski +8 more
TL;DR: NeGen as discussed by the authors combines an fMRI-trained neural encoding model of human vision with a deep generative network to synthesize images predicted to achieve a target pattern of macro-scale brain activation.
Proceedings ArticleDOI
Ensembling Low Precision Models for Binary Biomedical Image Segmentation
Tianyu Ma,Hang Zhang,Hanley Ong,Amar Vora,Thanh D. Nguyen,Ajay Gupta,Yi Wang,Mert R. Sabuncu +7 more
TL;DR: In this paper, a diverse ensemble of low precision and high recall models are trained to make different false positive errors (classifying background as foreground in different parts of the image), but the true positives will tend to be consistent.
Book ChapterDOI
Detecting Abnormalities in Resting-State Dynamics: An Unsupervised Learning Approach
TL;DR: In this article, an autoencoder approach on the rs-fMRI sequence and a next frame prediction strategy were explored for abnormality detection in the context of discriminating autism patients from healthy controls.
Journal ArticleDOI
UniverSeg: Universal Medical Image Segmentation
Victor Butoi,Jose Javier Gonzalez Ortiz,Tianyu Ma,Mert R. Sabuncu,John V. Guttag,Adrian V. Dalca +5 more
TL;DR: UniverSeg as discussed by the authors employs a cross-block mechanism to produce accurate segmentation maps without the need for additional training, which can be used for solving unseen medical segmentation tasks without additional training.
Posted Content
Learning the Distribution: A Unified Distillation Paradigm for Fast Uncertainty Estimation in Computer Vision
TL;DR: A unified distillation paradigm for learning the conditional predictive distribution of a pre-trained dropout model for fast uncertainty estimation of both aleatoric and epistemic uncertainty at the same time is proposed.