On multi-view feature learning

06/18/2012
by   Roland Memisevic, et al.
0

Sparse coding is a common approach to learning local features for object recognition. Recently, there has been an increasing interest in learning features from spatio-temporal, binocular, or other multi-observation data, where the goal is to encode the relationship between images rather than the content of a single image. We provide an analysis of multi-view feature learning, which shows that hidden variables encode transformations by detecting rotation angles in the eigenspaces shared among multiple image warps. Our analysis helps explain recent experimental results showing that transformation-specific features emerge when training complex cell models on videos. Our analysis also shows that transformation-invariant features can emerge as a by-product of learning representations of transformations.

READ FULL TEXT

page 6

page 7

research
02/13/2018

A probabilistic framework for multi-view feature learning with many-to-many associations via neural networks

A simple framework Probabilistic Multi-view Graph Embedding (PMvGE) is p...
research
06/27/2012

Learning Invariant Representations with Local Transformations

Learning invariant representations is an important problem in machine le...
research
04/21/2021

MVFuseNet: Improving End-to-End Object Detection and Motion Forecasting through Multi-View Fusion of LiDAR Data

In this work, we propose MVFuseNet, a novel end-to-end method for joint ...
research
01/13/2020

Learning Transformation-Aware Embeddings for Image Forensics

A dramatic rise in the flow of manipulated image content on the Internet...
research
08/10/2013

Learning Features and their Transformations by Spatial and Temporal Spherical Clustering

Learning features invariant to arbitrary transformations in the data is ...
research
12/08/2017

Transformational Sparse Coding

A fundamental problem faced by object recognition systems is that object...
research
12/20/2013

Unsupervised Feature Learning by Deep Sparse Coding

In this paper, we propose a new unsupervised feature learning framework,...

Please sign up or login with your details

Forgot password? Click here to reset