We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: Creativity Inspired Zero-Shot Learning

Abstract: Zero-shot learning (ZSL) aims at understanding unseen categories with no training examples from class-level descriptions. To improve the discriminative power of zero-shot learning, we model the visual learning process of unseen categories with inspiration from the psychology of human creativity for producing novel art. We relate ZSL to human creativity by observing that zero-shot learning is about recognizing the unseen and creativity is about creating a likable unseen. We introduce a learning signal inspired by creativity literature that explores the unseen space with hallucinated class-descriptions and encourages careful deviation of their visual feature generations from seen classes while allowing knowledge transfer from seen to unseen classes. Empirically, we show consistent improvement over the state of the art of several percents on the largest available benchmarks on the challenging task or generalized ZSL from a noisy text that we focus on, using the CUB and NABirds datasets. We also show the advantage of our approach on Attribute-based ZSL on three additional datasets (AwA2, aPY, and SUN). Code is available.
Comments: This paper was published at the International Conference on Computer Vision 2019, Seoul, South Korea, this http URL
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Journal reference: International Conference on Computer Vision-2019
Cite as: arXiv:1904.01109 [cs.CV]
  (or arXiv:1904.01109v6 [cs.CV] for this version)

Submission history

From: Mohamed Elhoseiny [view email]
[v1] Mon, 1 Apr 2019 21:05:23 GMT (2084kb,D)
[v2] Wed, 3 Apr 2019 17:18:51 GMT (2082kb,D)
[v3] Mon, 15 Apr 2019 21:43:56 GMT (3508kb,D)
[v4] Wed, 17 Apr 2019 02:57:29 GMT (3508kb,D)
[v5] Sun, 1 Dec 2019 19:27:06 GMT (3251kb,D)
[v6] Tue, 3 Dec 2019 02:42:29 GMT (3251kb,D)

Link back to: arXiv, form interface, contact.