scispace - formally typeset
R

Ross Goroshin

Researcher at Google

Publications -  27
Citations -  4565

Ross Goroshin is an academic researcher from Google. The author has contributed to research in topics: Image segmentation & Computer science. The author has an hindex of 14, co-authored 23 publications receiving 3703 citations. Previous affiliations of Ross Goroshin include Naval Surface Warfare Center & New York University.

Papers
More filters
Proceedings ArticleDOI

Efficient object localization using Convolutional Networks

TL;DR: In this paper, an efficient position refinement model is proposed to estimate the joint offset location within a small region of the image. And this model is jointly trained with a state-of-the-art ConvNet model to achieve improved accuracy in human joint location estimation.
Posted Content

Efficient Object Localization Using Convolutional Networks

TL;DR: A novel architecture which includes an efficient `position refinement' model that is trained to estimate the joint offset location within a small region of the image to achieve improved accuracy in human joint location estimation is introduced.
Proceedings Article

Learning to Navigate in Complex Environments

TL;DR: This work considers jointly learning the goal-driven reinforcement learning problem with auxiliary depth prediction and loop closure classification tasks and shows that data efficiency and task performance can be dramatically improved by relying on additional auxiliary tasks leveraging multimodal sensory inputs.
Journal ArticleDOI

Vector-based navigation using grid-like representations in artificial agents

TL;DR: These findings show that emergent grid-like representations furnish agents with a Euclidean spatial metric and associated vector operations, providing a foundation for proficient navigation, and support neuroscientific theories that see grid cells as critical for vector-based navigation.
Posted Content

Learning to Navigate in Complex Environments

TL;DR: In this paper, the authors formulate the navigation question as a reinforcement learning problem and show that data efficiency and task performance can be dramatically improved by relying on additional auxiliary tasks leveraging multimodal sensory inputs.