He is one of my favorite people in the world, funny, sweet, brilliant, as everyone already said. I just wanted to express my deepest condolences… this is very sad indeed. While at MIT and upon his return to Toronto, he focused on using machine learning and statistical methods to contribute to other sciences, such as astronomy and biology. I had the privilege of taking two courses with him, and even now find myself occasionally referring to his course notes. We worked closely for many years, we learned together, taught together and grew up academically together. His level of energy was fantastic — when we were climbing, skiing, jamming on problems, even when he was TAing , he could help push me to my limit like few others. The machine learning community has lost one of its best and most generous.
Thank you, John, for the post. Sam and David introduced astronomers to a number of large-scale statistical methods that enabled increasingly automated and precise data analysis. Sam was an amazingly generous friend and truly inspiring colleague. My great condolences to his family. Because of that, he is easily remembered and will be greatly missed. Let us be grateful that we knew him and his work. I discuss how these models can be used to learn a forward synthesis system which generates spectral sequences from articulatory movements.
Sam was a very close friend and I miss him deeply and painfully.
Sam Roweis, Associate Professor
Thank you for writing this. Unfortunately, despite potential rwoeis advantages and evidence for being part of the human strategy, such inversion of speech production models is absent in almost all artificial speech processing systems. He will be deeply missed. His first exposure to AI and neural computation occured when–as an exceptional undergraduate–he took the graduate-level Neural Network course taught by Geoffrey Hinton.
He proposed a variation of the well-established Hidden Markov Model HMM method for speech recognition, and he showed how a new form of Independent Component Analysis ICA could be used to separate multiple audio sources from a single microphone signal.
Despite his hectic schedule his door was always open to a wayward student seeking advice, academic or personal. Sam was a very unique person. I never met him, just knew him from his lectures online and his papers.
Data driven production models for speech processing
More information and software credits. I have long deeply admired Sam both personally and professionally. This permits organization, visualization, and search for large, complex data collections. It really is sad to lose such a great mind and, as many have echoed here, a great personality. They were born very prematurely and had to be kept in intensive care unit for many weeks. I just wanted to express my deepest condolences… this is very sad indeed.
There was such tremendous goodness in him. This gift allowed him to explain the key idea behind anything in just a few minutes. My last thewis with Sam, last week, was discussing a new research direction that interested him, optimizing the cost of acquiring feature information in the learning algorithm.
Take it easy now Sam. I wish I was able to roseis. Sam was a dear friend of our family and we are all so, so sad. He still has many friends and admirers here, and that room will now always seem emptier knowing his voice will never again fill it.
Obituary Page of Sam Roweis | NYU Computer Science
Citation Roweis, Sam T. I feel really lucky to have collaborated with him. I knew Sam as a teacher and a mentor as an undergraduate at Toronto.
It is funny how for at least some of us myself at once an instance and a recovering, I hope, instance of this it can truly feel like that to-do list consumes your life. His passing leaves many open threads, and many projects unfinished; at the time of his death he was working on beautiful, simple, but game-changing ideas for astronomical data rroweis and remote sensing. He was very approachable and friendly. I know that many of his friends will be glad to witness him in this professional capacity.
When difficult computations are to be performed on sensory data it is often advantageous to employ a model of the underlying process which produced the observations.
Sam inspired many students to pursue a career in research, and to focus their research on machine learning and artificial intelligence. Sam was my youngest cousin.
I never knew Sam academically and I sort of like it that way, he was my pal and friend. No commercial reproduction, distribution, display or performance rights in this work are provided. We worked closely for many years, we learned together, taught together and grew up academically together.
Sam had a singular gift: In making this choice Sam rejected several extremely prestigious offers for thesix unparalleled intellectual atmosphere he found at Toronto surrounding his mentor, Geoff.