mlsanityData to 7 Oct 2026

How saccadic vision might help with theinterpretability of deep networks

arXiv 2105.13264 ↗First posted 27 May 2021cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

We describe how some problems (interpretability,lack of object-orientedness) of modern deep networks potentiallycould be solved by adapting a biologically plausible saccadicmechanism of perception. A sketch of such a saccadic visionmodel is proposed. Proof of concept experimental results areprovided to support the proposed approach.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 15 of its 29 references that are in the map. First posted on arXiv 27 May 2021.

Builds on

15 of 29 references.

References · 15

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  3. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 5,986
  4. 2016Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotationscited by 2,646
  5. 2017A systematic study of the class imbalance problem in convolutional neural networkscited by 510
  6. 2017Scene Graph Generation by Iterative Message Passingcited by 462
  7. 2019Making Convolutional Networks Shift-Invariant Againcited by 390
  8. 2019AutoML: A Survey of the State-of-the-Artcited by 388
  9. 2017Measuring Catastrophic Forgetting in Neural Networkscited by 359
  10. 2014Recurrent Models of Visual Attentioncited by 344
  11. 2018An Intriguing Failing of Convolutional Neural Networks and the CoordConv Solutioncited by 302
  12. 2018Graph R-CNN for Scene Graph Generationcited by 296
3 more
  1. 2020Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Futurecited by 83
  2. 2019Differentiable Scene Graphscited by 16
  3. 2020Problems of representation of electrocardiograms in convolutional neural networkscited by 1

Built on by

Not cited yet.