Testing Deep Neural Networks on the Same-Different Task

@article{Messina2019TestingDN,
  title={Testing Deep Neural Networks on the Same-Different Task},
  author={Nicola Messina and Giuseppe Amato and Fabio Carrara and F. Falchi and Claudio Gennaro},
  journal={2019 International Conference on Content-Based Multimedia Indexing (CBMI)},
  year={2019},
  pages={1-6}
}
  • Nicola Messina, G. Amato, +2 authors C. Gennaro
  • Published 1 September 2019
  • Computer Science
  • 2019 International Conference on Content-Based Multimedia Indexing (CBMI)
Developing abstract reasoning abilities in neural networks is an important goal towards the achievement of human-like performances on many tasks. As of now, some works have tackled this problem, developing ad-hoc architectures and reaching overall good generalization performances. In this work we try to understand to what extent state-of-the-art convolutional neural networks for image classification are able to deal with a challenging abstract problem, the so-called same-different task. This… Expand
Solving the Same-Different Task with Convolutional Neural Networks
TLDR
This work probes current state-of-the-art convolutional neural networks on a difficult set of tasks known as the same-different problems, and shows that older feed-forward networks, AlexNet and VGG, are almost unable to learn the proposed problems, except in some specific scenarios. Expand
The Notorious Difficulty of Comparing Human and Machine Perception
TLDR
It is shown that, despite their ability to solve closed-contour tasks, the authors' neural networks use different decision-making strategies than humans, and that neural networks do experience a "recognition gap" on minimal recognizable images. Expand
Combining EfficientNet and Vision Transformers for Video Deepfake Detection
TLDR
This study combines various types of Vision Transformers with a convolutional EfficientNet B0 used as a feature extractor, obtaining comparable results with some very recent methods that use Vision Transformers. Expand
Arguments for the Unsuitability of Convolutional Neural Networks for Non-Local Tasks
TLDR
This paper will use findings on sorting networks to propose a lower bound for the number of parameters a neural network needs to solve comparison tasks in a generalizable way, and argue that attention, as well as iterative/recurrent processing, is needed to prevent a combinatorial explosion. Expand
Evaluating the progress of deep learning for visual relational concepts
TLDR
It will be hypothesised that iterative processing of the input, together with shifting attention between the iterations will be needed to efficiently and reliably solve real world relational concept learning. Expand
Five points to check when comparing visual perception in humans and machines
TLDR
A checklist for comparative studies of visual reasoning in humans and machines is presented and it is found that a previously observed difference in object recognition does not hold when adapting the experiment to make conditions more equitable between humans and machine. Expand

References

SHOWING 1-10 OF 23 REFERENCES
A dataset and architecture for visual reasoning with a working memory
TLDR
This work developed an artificial, configurable visual question and answer dataset (COG) to parallel experiments in humans and animals and proposes a deep learning architecture that performs competitively on other diagnostic VQA datasets (i.e. CLEVR) as well as easy settings of the COG dataset. Expand
Going deeper with convolutions
We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual RecognitionExpand
Not-So-CLEVR: Visual Relations Strain Feedforward Neural Networks
TLDR
Motivated by the comparable success of biological vision, it is argued that feedback mechanisms including working memory and attention are the key computational components underlying abstract visual reasoning. Expand
A simple neural network module for relational reasoning
TLDR
This work shows how a deep learning architecture equipped with an RN module can implicitly discover and learn to reason about entities and their relations. Expand
25 Years of CNNs: Can We Compare to Human Abstraction Capabilities?
TLDR
The performance of LeNet is compared to that of GoogLeNet at classifying randomly generated images which are differentiated by an abstract property to show that there is still work to do in order to solve vision problems humans are able to solve without much difficulty. Expand
ImageNet classification with deep convolutional neural networks
TLDR
A large, deep convolutional neural network was trained to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes and employed a recently developed regularization method called "dropout" that proved to be very effective. Expand
Measuring abstract reasoning in neural networks
TLDR
A dataset and challenge designed to probe abstract reasoning, inspired by a well-known human IQ test, is proposed and ways to both measure and induce stronger abstract reasoning in neural networks are introduced. Expand
CORnet: Modeling the Neural Mechanisms of Core Object Recognition
TLDR
The current best ANN model derived from this approach (CORnet-S) is among the top models on Brain-Score, a composite benchmark for comparing models to the brain, but is simpler than other deep ANNs in terms of the number of convolutions performed along the longest path of information processing in the model. Expand
Very deep convolutional neural network based image classification using small training sample size
TLDR
The results show that the very deep CNN can be used to fit small datasets with simple and proper modifications and don't need to re-design specific small networks. Expand
Deep Residual Learning for Image Recognition
TLDR
This work presents a residual learning framework to ease the training of networks that are substantially deeper than those used previously, and provides comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. Expand
...
1
2
3
...