CaltechAUTHORS
  A Caltech Library Service

Task2Vec: Task Embedding for Meta-Learning

Achille, Alessandro and Lam, Michael and Tewari, Rahul and Ravichandran, Avinash and Maji, Subhransu and Fowlkes, Charless and Soatto, Stefano and Perona, Pietro (2019) Task2Vec: Task Embedding for Meta-Learning. In: 2019 IEEE/CVF International Conference on Computer Vision (ICCV). IEEE , Piscataway, NJ, pp. 6429-6438. ISBN 9781728148038. https://resolver.caltech.edu/CaltechAUTHORS:20190327-085927491

[img] PDF - Submitted Version
See Usage Policy.

1553Kb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20190327-085927491

Abstract

We introduce a method to generate vectorial representations of visual classification tasks which can be used to reason about the nature of those tasks and their relations. Given a dataset with ground-truth labels and a loss function, we process images through a "probe network" and compute an embedding based on estimates of the Fisher information matrix associated with the probe network parameters. This provides a fixed-dimensional embedding of the task that is independent of details such as the number of classes and requires no understanding of the class label semantics. We demonstrate that this embedding is capable of predicting task similarities that match our intuition about semantic and taxonomic relations between different visual tasks. We demonstrate the practical value of this framework for the meta-task of selecting a pre-trained feature extractor for a novel task. We present a simple meta-learning framework for learning a metric on embeddings that is capable of predicting which feature extractors will perform well on which task. Selecting a feature extractor with task embedding yields performance close to the best available feature extractor, with substantially less computational effort than exhaustively training and evaluating all available models.


Item Type:Book Section
Related URLs:
URLURL TypeDescription
https://doi.org/10.1109/iccv.2019.00653DOIArticle
https://arxiv.org/abs/1902.03545arXivDiscussion Paper
ORCID:
AuthorORCID
Fowlkes, Charless0000-0002-2990-1780
Perona, Pietro0000-0002-7583-5809
Additional Information:© 2019 IEEE.
Record Number:CaltechAUTHORS:20190327-085927491
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20190327-085927491
Official Citation:A. Achille et al., "Task2Vec: Task Embedding for Meta-Learning," 2019 IEEE/CVF International Conference on Computer Vision (ICCV), Seoul, Korea (South), 2019, pp. 6429-6438. doi: 10.1109/ICCV.2019.00653
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:94203
Collection:CaltechAUTHORS
Deposited By: George Porter
Deposited On:27 Mar 2019 16:21
Last Modified:06 Mar 2020 22:37

Repository Staff Only: item control page