Technical Talks- Afternoon

2020 Afternoon Technical Talks

Dr. Andrea Arnold

Assistant professor, Mathematical Science, Bioinformatics and Computational Biology, WPI

Dr. Andrea Arnold is an Assistant Professor in the Department of Mathematical Sciences at WPI.  Her research is in the field of inverse problems and uncertainty quantification, which involves estimating unknown system parameters using indirect observations and analyzing the changes in predicted outcomes due to changes in the inputs.  Dr. Arnold’s work focuses in particular on the design and analysis of efficient and robust nonlinear filtering algorithms for state and parameter estimation.  She applies these algorithms in analyzing real-world data from areas of the life sciences and engineering, with a specific interest in patient-specific models for biomedicine.  Dr. Arnold is the faculty advisor for the WPI Student Chapter of the Association for Women in Mathematics (AWM).


Talk title: TBA

Abstract: TBA


Dr. Esra Cansizoglu

Machine Learning Engineer, Facebook

Esra Cansizoglu has been working on machine learning and computer vision problems in various domains: from understanding interior design to object localization for robotic perception. She received his PhD in Electrical Engineering from Northeastern University and MS in Computer Science from Boston University. She worked at Mitsubishi Electric Research Labs as a research scientist. Later she joined computer vision team at Wayfair and worked on image-based models for product recommendation. She recently joined Facebook Boston office as a machine learning engineer.

Talk title : Deep Ranking for Style-Aware Room Recommendations

Abstract: We present a deep learning based room image retrieval framework that is based on style understanding. Given a dataset of room images labeled by interior design experts, we map the noisy style labels to comparison labels. Our framework learns the style spectrum of each image from the generated comparisons and makes significantly more accurate recommendations compared to discrete classification baselines.