Robust Object Recognition with Cortex-Like Mechanisms

By Thomas Serre, Lior Wolf, S. Bileschi, Maximilian Riesenhuber, Tomaso A. Poggio. 2007

1,855 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Biophysics, 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 Artificial Intelligence

In Biophysics

In Cognitive Neuroscience

In Computer Vision and Pattern Recognition

Citations in the corpus

Going deeper with convolutions (2014, methodology): “Inspired by a neuroscience model of the primate visual cortex, Serre et al. [15] use a series of ﬁxed Gabor ﬁlters of different sizes in order to handle multiple scales, similarly to the Inception model.”
Representation Learning: A Review and New Perspectives (2012, background): “…pooling) has been the basis of convolutional networks, the Neocognitron (Fukushima, 1980) and HMAX (Riesenhuber and Poggio, 1999) models, and argued to be the architecture used by mammalian brains for object recognition (Riesenhuber and Poggio, 1999; Serre et al., 2007; DiCarlo et al., 2012).”
Multi-column deep neural networks for image classification (2012, background): “such as unsupervised pre-training [29, 24, 2, 10] or carefully prewired synapses [27, 31].”
HMDB: A large video database for human motion recognition (2011, methodology): “Te width was scaled accordingly to maintain the original aspect ratio.”
View-tolerant face recognition and Hebbian learning imply mirror-symmetric neural tuning to head orientation (2016, positive): “Current computational models of object recognition, including recent deep learning networks, generate these properties through a hierarchy of alternating selectivity-increasing filtering and tolerance-increasing pooling operations, similar to simple-complex cells operations [46, 8, 44, 29].”
Unmasking Clever Hans predictors and assessing what machines really learn (2019, background): “edges → corners → rectangles → boxes → cars) mirroring abstraction levels similar to the human brain [107, 108].”
Generalizing Pooling Functions in Convolutional Neural Networks: Mixed, Gated, and Tree (2015, background): “The recent resurgence of neurally-inspired systems such as deep belief nets (DBN) [10], convolutional neural networks (CNNs) [18], and the sum-and-max infrastructure [32] has derived significant benefit from building more sophisticated network structures [38, 33] and from bringing learning to non-li”
How does the brain solve visual object recognition? (2012, positive): “This can produce a gradual increase of the tolerance to variation in object appearance along the hierarchy (e.g., Fukushima, 1980; Riesenhuber and Poggio, 1999b; Serre et al., 2007a).”
Distributed optimization of deeply nested systems (2012, background): “Other examples of hierarchical processing systems are cascades for object recognition and scene understanding in computer vision (Serre et al., 2007; Ranzato et al., 2007a) or for phoneme classiﬁcation in speech processing (Gold and Morgan, 2000; Saon and Chien, 2012), wrapper approaches to…”
A High-Throughput Screening Approach to Discovering Good Forms of Biologically Inspired Visual Representation (2009, neutral): “Of note, the nearest contender from the set of state-of-the-art models is another biologicallyinspired model [7,8].”
A Biologically Inspired System for Action Recognition (2007, methodology): “Learning sparse spatio-temporal motion S2 features In the work [61], a Gaussian-like function is used to compute the responses of dense S2 features.”
From the Phenomenology to the Mechanisms of Consciousness: Integrated Information Theory 3.0 (2014, background): “Moreover, recent advances in machine learning have made it possible to construct simulated networks, primarily feed-forward, that can learn to recognize natural categories such as cats, dogs [1], pedestrians [56,57], and/or faces [58–60].”
Learning Hybrid Image Templates (HIT) by Information Projection (2012, methodology): “The local maximum pooling is proposed by [33] as a possible function of complex cells in V1.”
Space-Variant Descriptor Sampling for Action Recognition Based on Saliency and Eye Movements (2012, background): “Saliency-based approaches have been successfully applied to various computer vision tasks, too, such as image compression [10], quality assessment, and object recognition [11,12,13].”
One-shot Texture Segmentation (2018, methodology): “In [35] a biologically motivated model of texture perception has been used to recognize objects.”
Why is Real-World Visual Object Recognition Hard? (2008, background): “However, annotating such an image set is extremely labor-intensive (but see the LabelMe project [25], Peekaboom [26], and the StreetScenes dataset [2,27]).”