A
Andrew Rabinovich
Researcher at Google
Publications - 67
Citations - 51886
Andrew Rabinovich is an academic researcher from Google. The author has contributed to research in topics: Convolutional neural network & Artificial neural network. The author has an hindex of 28, co-authored 67 publications receiving 37872 citations. Previous affiliations of Andrew Rabinovich include University of California, San Diego & Discovery Institute.
Papers
More filters
Patent
Selection of representative images
Anand Pillai,Andrew Rabinovich +1 more
TL;DR: In this paper, a computer-implemented method for selecting a representative image of an entity is disclosed, which includes: accessing a collection of images of the entity; clustering, based on similarity of one or more similarity features, images from the collection to form a plurality of similarity clusters; and selecting the representative image from one of said similarity clusters.
Posted Content
Self-Improving Visual Odometry
TL;DR: A self-supervised learning framework that uses unlabeled monocular video sequences to generate large-scale supervision for training a Visual Odometry frontend, a network which computes pointwise data associations across images.
Patent
Gradient adversarial training of neural networks
TL;DR: gradient adversarial training increases the robustness of a network to adversarial attacks, is able to better distill the knowledge from a teacher network to a student network compared to soft targets, and boosts multi-task learning by aligning the gradient tensors derived from the task specific loss functions.
Posted Content
Deep Cuboid Detection: Beyond 2D Bounding Boxes
TL;DR: This work proposes an end-to-end deep learning system to detect cuboids across many semantic categories, and localizes all 3D cuboids (box-like objects) with a 2D bounding box.
Posted Content
RoomNet: End-to-End Room Layout Estimation
TL;DR: In this paper, an end-to-end trainable encoder-decoder network is proposed to predict the locations of the room layout keypoints, which achieves state-of-the-art performance on the challenging benchmark datasets Hedau and LSUN.