Unsupervised Learning using Pretrained CNN and Associative Memory Bank

2 May 2018  ·  Qun Liu, Supratik Mukhopadhyay ·

Deep Convolutional features extracted from a comprehensive labeled dataset, contain substantial representations which could be effectively used in a new domain. Despite the fact that generic features achieved good results in many visual tasks, fine-tuning is required for pretrained deep CNN models to be more effective and provide state-of-the-art performance. Fine tuning using the backpropagation algorithm in a supervised setting, is a time and resource consuming process. In this paper, we present a new architecture and an approach for unsupervised object recognition that addresses the above mentioned problem with fine tuning associated with pretrained CNN-based supervised deep learning approaches while allowing automated feature extraction. Unlike existing works, our approach is applicable to general object recognition tasks. It uses a pretrained (on a related domain) CNN model for automated feature extraction pipelined with a Hopfield network based associative memory bank for storing patterns for classification purposes. The use of associative memory bank in our framework allows eliminating backpropagation while providing competitive performance on an unseen dataset.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Fine-Grained Image Classification Caltech-101 UL-Hopfield (ULH) Accuracy 91.00% # 1
Semi-Supervised Image Classification Caltech-101 UL-Hopfield (ULH) Accuracy 91.00% # 1
Semi-Supervised Image Classification Caltech-101, 202 Labels UL-Hopfield (ULH) Accuracy 91.00% # 1
Semi-Supervised Image Classification Caltech-256 UL-Hopfield (ULH) Accuracy 77.40% # 1
Semi-Supervised Image Classification Caltech-256, 1024 Labels UL-Hopfield (ULH) Accuracy 77.40% # 1
Few-Shot Image Classification Caltech-256 5-way (1-shot) UL-Hopfield (ULH) Accuracy 74.7 # 1
Image Classification CIFAR-10 UL-Hopfield (ULH) Percentage correct 83.1 # 212
Few-Shot Image Classification CIFAR100 5-way (1-shot) UL-Hopfield (ULH) Accuracy 89.6 # 1
Semi-Supervised Image Classification CIFAR-10, 40 Labels UL-Hopfield (ULH) Percentage error 16.90 # 17

Methods


No methods listed for this paper. Add relevant methods here