Machine Learning

Eric Cosatto

Senior Researcher

Machine Learning

NEC Laboratories America, Inc.

4 Independence Way, Suite 200
Princeton, NJ 08540

Telephone: +1 609-951-4811

Fax: +1 609-951-2482

Email: cosatto@nec-labs.com

Publications

Automatic Fine-Grained Localization of Utility Pole Landmarks on Distributed Acoustic Sensing Traces Based on Bilinear Resnets
2021 IEEE International Conference on Acoustics, Speech and Signal Processing
pp. 4675-4679, 2021
You Lu, Yue Tian, Shaobo Han, Eric Cosatto, Sarper Ozharar, Yangmin Ding
06/06/2021
A Multi-Scale Conditional Deep Model for Tumor Cell Ratio Counting
SPIE Medical Imaging 2021: MI109: Digital and Computational Pathology
11603:1160308-1-1160308-12, 2021
Eric Cosatto, Kyle Gerard, Hans-Peter Graf, Maki Ogura, Tomoharu Kiyuna, Kanako C Hatanaka, Yoshihiro Matsuno, Yutaka Hatanaka
02/17/2021
Prediction of Early Recurrence of Hepatocellular Carcinoma after Resection using Digital Pathology Images Assessed by Machine Learning
Modern Pathology
34:417-425, 2021
Akira Saito, Hidenori Toyoda, Masaharu Kobayashi, Yoshinori Koiwa, Hiroki Fujii, Koji Fujita, Atsuyuki Maeda, Yuji Kaneoka, Shoichi Hazama, Hiroaki Nagano, Aashiq H. Mirza, Hans-Peter Graf, Eric Cosatto, Yoshiki Murakami, Masahiko Kuroda
09/18/2020

More Publications

Projects

Digital Pathology

Accurate and fast pathological diagnosis is crucial to prevention, early detection and early treatment in defeating cander. NEC is developing a cancer diagnosis assistance system using digital image processing technologies and machine learning algorithms. With this system, NEC aims to help improve the administrative efficiency and diagnostic quality of pathological diagnosis READ MORE

Parallel Computation in Learning

We explore algorithms for implementing large scale learning algorithms as parallel computation. We are currently developing parallelization approaches for increasing the ability of SVM (Support Vector Machines) to solve large-scale problems. As target systems, we consider shared memory processors, clusters of processors, vector processors, and SIMD (Single Instruction Multiple Data) processors. On a given system the speed of an SVM is limited by the computer performance of the processor as well as by the size of the memory. Efficient parallelizations have to overcome both of these limitations while not getting bogged down in communication overhead. READ MORE

Software

MiLDe

MiLDe is an integrated development environment with a suite of machine learning tools including SVMs, Neural Networks (using the Torch package), image analysis, numeric functions and dataset handling. READ MORE

Eric Cosatto

Publications

Automatic Fine-Grained Localization of Utility Pole Landmarks on Distributed Acoustic Sensing Traces Based on Bilinear Resnets

A Multi-Scale Conditional Deep Model for Tumor Cell Ratio Counting

Prediction of Early Recurrence of Hepatocellular Carcinoma after Resection using Digital Pathology Images Assessed by Machine Learning

Princeton Campus

San Jose Campus