Deep Neural Networks Point to Mid-level Complexity of Rodent Object Vision

Abstract

1In the last two decades rodents have been on the rise as a dominant model for visual 2 neuroscience. This is particularly true for earlier levels of information processing, but 3 high-profile papers have suggested that also higher levels of processing such as invariant 4 object recognition occur in rodents. Here we provide a quantitative and comprehensive 5 assessment of this claim by comparing a wide range of rodent behavioral and neural data 6 with convolutional deep neural networks. These networks have been shown to capture 7 the richness of information processing in primates through a succession of convolutional 8 and fully connected layers. We find that rodent object vision can be captured using 9 low to mid-level convolutional layers only, without any convincing evidence for the 10 need of higher layers known to simulate complex object recognition in primates. Our 11 approach also reveals surprising insights on assumptions made before, for example, that 12 the best performing animals would be the ones using the most complex representations 13 -which we show to likely be incorrect. Our findings suggest a road ahead for further 14 studies aiming at quantifying and establishing the richness of representations underlying 15 information processing in animal models at large. 16 1 Introduction 17 Up to one decade ago, macaque monkeys were uncontested as the primary animal model for 18 vision research targeting information processing at the cortical level. Since then, studies on 19 rodents have become increasingly popular, in particular because developments in neurotech-20 nology such as cell-level imaging and optogenetics capitalized upon the existence of genetic 21 models. A major question emerging from this evolution is how far we can take rodents as a 22 model for the more complex aspects of visual information processing. 23To answer this question, a series of high-profile studies have documented the ability of rats 24 to perform seemingly complex object recognition tasks. In these tasks, rats show the ability 25 to recognize objects despite various transformations in viewing conditions such as position, 26 size, and viewpoint. In the first landmark study, Zoccolan and colleagues (2009) trained rats 27 2 to discriminate two computer-graphics rendered objects under various changes in size and 28 rotation, and showed that the animals could successfully generalize to novel combinations 29 of these transformations. Several studies have subsequently expanded on this work using 30 similar stimuli (Tafazoli et al., 2012; Alemi-Neissi et al., 2013; Rosselli et al., 2015), and 31 recently it was found that the success of rats in these tasks depends on the complexity of 32 their strategy (Djurdjevic et al., 2018). In a study using natural stimuli, rats could generalize 33 learned category rules to new, unseen category exemplars that differ from trained exemplars 34 in complex ways (Vinken et al., 2014). Neurophysiological recordings have revealed neural 35 responses in lateral visual areas that might underlie these b...

Neural Regression, Representational Similarity, Model Zoology & Neural Taskonomy at Scale in Rodent Visual Cortex

“…Echoing previous results [7,8], we find across all ImageNet-trained architectures, regardless of metric, that the features most predictive of rodent visual cortex are found about a third of the way into the model (though see Section A.5 of the Appendix for some caveats). These early to intermediate visual features go beyond basic edge detection but are far from the highly abstracted representations adjacent to final fully connected layers.…”

“…Corresponding to the biology not only at the level of individual layers, but across the feature hierarchy, these models are so powerful they can now effectively be used as neural controllers, synthesizing stimuli that drive neural activity far beyond the range evoked by any handmade experimental stimulus [6]. The correspondence of these same models to mouse visual cortex, on the other hand, has proven a bit more tenuous [7, 8], with a recent finding even suggesting that randomly initialized networks are as predictive of rodent visual cortical activity as trained ones [9].…”

Mice and primates use distinct strategies for visual segmentation

“…Our finding of strong texturedependence in mouse segmentation behavior suggests that mice adopt a visual strategy more similar to deep networks than primates do (Fig. 7g; see also [55]).…”

Brain organization, not size alone, as key to high-level vision: Evidence from marmoset monkeys

“…This result demonstrates that high performance on the rodent task does not require sophisticated representations for visual form processing. Therefore, despite previous interpretations 30 , this paradigm may not be a strong behavioral test of invariant object recognition (also see 32 ). Even on this relatively simple visual task, marmosets still out-performed rodents and generalized in a far more robust manner.…”