25 March 2022 in IEEE Transactions on Neural Networks and Learning Systems
The ResNet and its variants have achieved remarkable successes in various computer vision tasks. Despite its success in making gradient flow through building blocks, the information communication of intermediate layers of blocks is ignored. To address this issue, in this brief, we propose to introduce a regulator module as a memory mechanism to extract complementary features of the intermediate layers, which are further fed to the ResNet. In particular, the regulator module is composed of convolutional recurrent neural networks (RNNs) [e.g., convolutional long short-term memories (LSTMs) or convolutional gated recurrent units (GRUs)], which are shown to be good at extracting spatio-temporal information. We named the new regulated network as regulated residual network (RegNet). The regulator module can be easily implemented and appended to any ResNet architecture. Experimental results on three image classification datasets have demonstrated the promising performance of the proposed architecture compared with the standard ResNet, squeeze-and-excitation ResNet, and other state-of-the-art architectures.
23 February 2024 in IEEE Transactions on Neural Networks and Learning Systems
Few-shot classification aims to adapt classifiers trained on base classes to novel classes with a few shots. However, the limited amount of training data is often inadequate to represent the intraclass variations in novel classes. This can result in biased estimation of the feature distribution, which in turn results in inaccurate decision boundaries, especially when the support data are outliers. To address this issue, we propose a feature enhancement method called CORrelation-guided feature Enrichment that generates improved features for novel classes using weak supervision from the base classes. The proposed CORrelation-guided feature Enhancement (CORE) method utilizes an autoencoder (AE) architecture but incorporates classification information into its latent space. This design allows the CORE to generate more discriminative features while discarding irrelevant content information. After being trained on base classes, CORE’s generative ability can be transferred to novel classes that are similar to those in the base classes. By using these generative features, we can reduce the estimation bias of the class distribution, which makes few-shot learning (FSL) less sensitive to the selection of support data. Our method is generic and flexible and can be used with any feature extractor and classifier. It can be easily integrated into existing FSL approaches. Experiments with different backbones and classifiers show that our proposed method consistently outperforms existing methods on various widely used benchmarks.
1 September 2013 in Advanced Materials Research
Under high temperature condition, determination of phase transition temperature and combination of PCM layer with the other layers affect the thermal buffering properties of the protective clothing. In this paper, Body-clothing-environment temperature gradient was analyzed. Related data was obtained from the dressing experiments. After that, phase transition temperature range of the protective clothing was determined. Temperature and humidity of human skin and microclimate were tested respectively with different positions of PCM layer. Finally, the optimization layer configuration was suggested.
10 November 2022 in arXiv
Standard language model training employs gold human documents or human-human interaction data, and treats all training data as positive examples. Growing evidence shows that even with very large amounts of positive training data, issues remain that can be alleviated with relatively small amounts of negative data -- examples of what the model should not do. In this work, we propose a novel procedure to train with such data called the CRINGE loss (ContRastive Iterative Negative GEneration). We show the effectiveness of this approach across three different experiments on the tasks of safe generation, contradiction avoidance, and open-domain dialogue. Our models outperform multiple strong baselines and are conceptually simple, easy to train and implement.