About
Rene Vidal received his B.S. degree in Electrical Engineering (highest honors) from the…
Contributions
-
Here's how you can effectively prioritize machine learning tasks and projects with input from your boss.
I agree, but this issue is not new. In the early 2000 computer vision researchers we were concerned with the evaluation of object detection systems: is it the same to hallucinate a pedestrian on a street vs hallucinating a pedestrian in the middle of the sky? Twenty years have passed and we have not fixed the issue.
Activity
-
Check out this new paper for Video Generation. "VideoPoet: A large language model for zero-shot video generation." Presented at the ICML 2024, in…
Check out this new paper for Video Generation. "VideoPoet: A large language model for zero-shot video generation." Presented at the ICML 2024, in…
Liked by Rene Vidal
-
Bonjour LinkedIn ! 👋 C'est avec enthousiasme que nous lançons aujourd'hui la présence officielle de l'Agence Ministérielle pour l'IA de Défense…
Bonjour LinkedIn ! 👋 C'est avec enthousiasme que nous lançons aujourd'hui la présence officielle de l'Agence Ministérielle pour l'IA de Défense…
Liked by Rene Vidal
-
It was great to be part of the multimodal-representations panel discussion at SIGIR 2024 Conference with Prof. Dinesh Manocha, Prof. Hamed Zamani and…
It was great to be part of the multimodal-representations panel discussion at SIGIR 2024 Conference with Prof. Dinesh Manocha, Prof. Hamed Zamani and…
Liked by Rene Vidal
Experience
Education
-
University of California, Berkeley
-
Activities and Societies: International House
Developed vision-based algorithms for landing a helicopter on a moving platform.
-
-
Activities and Societies: Salsa dancing
Studied multibody multiview geometry and developed motion segmentation algorithms
-
-
Studied electrical engineering.
-
-
Studied electrical engineering
Publications
-
Global Optimality in Tensor Factorization, Deep Learning, and Beyond
arxiv
While deep learning approaches have revolutionized machine learning, computer vision, speech and language processing, the theoretical reasons for its success remain elusive. This work studies conditions for global optimality in deep learning. In particular, it is shown that if the output of the network and the network regularizer are positively homogeneous functions of the network weights, all local minima that satisfy a simple condition are global minima.
Other authors -
Patents
-
System and method for aligning video sequences
Issued US US 9025908 B2
Methods and Systems for aligning multiple video sequences of a similar scene. It is determined which video sequences should be aligned with each other using linear dynamic system (LDS) modeling. The video sequences are then spatially aligned with each other.
Projects
-
Sparse Subspace Clustering
- Present
Subspace clustering is an important problem with numerous applications in image processing and computer vision. Given a set of points drawn from a union of linear or affine subspaces, the task is to find segmentation of the data. In most applications the data are embedded in high-dimensional spaces, while the underlying subspaces are low-dimensional. In this project, we propose a new approach to subspace clustering based on sparse representation. We exploit the fact that each data point in a…
Subspace clustering is an important problem with numerous applications in image processing and computer vision. Given a set of points drawn from a union of linear or affine subspaces, the task is to find segmentation of the data. In most applications the data are embedded in high-dimensional spaces, while the underlying subspaces are low-dimensional. In this project, we propose a new approach to subspace clustering based on sparse representation. We exploit the fact that each data point in a union of subspaces can always be written as a linear or affine combination of all other points. By searching for the sparsest combination, we automatically obtain a representation from points lying in the same subspace. This allows us to build a similarity matrix, from which the segmentation of the data can be easily obtained using spectral clustering. While in principle finding the sparsest representation is an NP hard problem, we show that under mild assumptions on the distribution of the data across subspaces, the sparsest representation can be found efficiently by solving a convex optimization problem.
Other creators -
-
GPCA
-
GPCA is an algebraic algorithm for clustering data drawn from a union of subspaces.
Other creators -
Honors & Awards
-
IAPR Fellow
IAPR
For contributions to computer vision and pattern recognition.
-
IEEE Fellow
IEEE
-
Best Paper Award ICCV Workshop on 3D Reconstruction and Recognition
IEEE
-
Best Paper Award, ICCV Workshop on 3D Representation and Recognition
-
-
Best Paper Award, Conference on Decision and Control
-
-
J.K. Aggarwal Award Prize
-
-
Best Paper Award in Medical Robotics and Computer Assisted Interventions
MICCAI
Best Paper Award in Medical Robotics and Computer Assisted Interventions for paper entitled “Surgical Gesture Classification from Video Data”, MICCAI 2012
Languages
-
English
-
-
Spanish
-
-
French
-
Organizations
-
IEEE
Associate Editor of TPAMI
- Present
More activity by Rene
-
The premiere AI meeting in Latinamerica is coming to Chile
The premiere AI meeting in Latinamerica is coming to Chile
Shared by Rene Vidal
-
"Standing on the Shoulders of Giants" Recently, I had the opportunity to revisit the place where I spent nearly six transformative years completing…
"Standing on the Shoulders of Giants" Recently, I had the opportunity to revisit the place where I spent nearly six transformative years completing…
Liked by Rene Vidal
-
l don't usually post to LinkedIn, but I wanted to thank so much my colleagues, whose collaborations and inspiration over the years made it possible…
l don't usually post to LinkedIn, but I wanted to thank so much my colleagues, whose collaborations and inspiration over the years made it possible…
Liked by Rene Vidal
-
Joyful scenes unfold at the 2024 Doctoral Commencement. Congratulations to all the new Penn Engineering graduates! #PennEngineeringProud
Joyful scenes unfold at the 2024 Doctoral Commencement. Congratulations to all the new Penn Engineering graduates! #PennEngineeringProud
Liked by Rene Vidal
-
The #nvidia $NVDA bull case explained to kids Source: Jesse Cohen @JesseCohenInv
The #nvidia $NVDA bull case explained to kids Source: Jesse Cohen @JesseCohenInv
Liked by Rene Vidal
-
Exciting news within Penn's innovation ecosystem! Penn Engineering launched its online Master of Science in Engineering (MSE) in Artificial…
Exciting news within Penn's innovation ecosystem! Penn Engineering launched its online Master of Science in Engineering (MSE) in Artificial…
Liked by Rene Vidal
-
Amazing to see LLMs and 3D Computer Vision coming together here!
Amazing to see LLMs and 3D Computer Vision coming together here!
Liked by Rene Vidal
-
Ayer tuve el honor de tomar posesión como nuevo Académico en la Real Academia de Farmacia de Galicia. Este reconocimiento es especialmente…
Ayer tuve el honor de tomar posesión como nuevo Académico en la Real Academia de Farmacia de Galicia. Este reconocimiento es especialmente…
Liked by Rene Vidal
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Rene Vidal in United States
-
René Vidal
7x NCAA Championship Coach | Head Coach, NIU Men's Tennis | 2022 GLVC Coach of the Year | Author + Podcast Host at Down 40 Love | Speaker on leadership, culture, & the psychology of winning.
-
Rene Vidal
CEO at ServiceMaster Commercial Cleaning by the Experts
-
Rene Vidal
Business administration majoring in International Business and Finance
-
Rene Vidal
Protocol Officer at UAE Embassy Washington DC - Military Officer
28 others named Rene Vidal in United States are on LinkedIn
See others named Rene Vidal