Humans, but Not Deep Neural Networks, Often Miss Giant Targets in Scenes.
By Miguel P. Eckstein, Kathryn Koehler, Lauren E. Welbourne, Emre Akbas. 2017
80 citations recorded by the source corpus.
Subfields: Cognitive Neuroscience, Computer Vision and Pattern Recognition
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Cognitive Neuroscience
- Matthias Bethge
- Robert Geirhos
- Wieland Brendel
- Felix Wichmann
- Claudio Michaelis
- Daniel Zoran
- Patrick Esser
- Axel Sauer
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Matthias Bethge
- Trevor Darrell
- Ming-Hsuan Yang
- Andrea Vedaldi
- Bernt Schiele
- Dacheng Tao
- Jitendra Malik
Citations in the corpus
- ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness (2018, background): “, 2018) and object size relative to the context (Eckstein et al., 2017) do not have a strong detrimental impact on CNN recognition performance, this highlights the special role that local cues such as textures seem to play in CNN object recognition.”
- Towards Robust Image Classification Using Sequential Attention Models (2019, background): “This has been postulated to cause humans to report different classification mistakes which are qualitatively different than those of deep neural networks [13].”