Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems

Software defect prediction (SDP) exerts a major role in software development, concerning reducing software costs and ensuring software quality. However, developing an accurate SDP model is still a severe and challenging task with the lack of training data. Fortunately, Siamese networks are powerful for learning a few samples and have been perfectly used in other fields. This paper explores the advantages of Siamese networks to propose a novel SDP model, Siamese dense neural networks (SDNNs), which integrates similarity feature learning and distance metric learning into a unified approach. It mainly includes two phases: model building and training. To be more specific, it means building the novel SDNN for capturing the highest-level similarity features and training the model to realize prediction through the designed contrast loss function with cosine proximity. Importantly, we extensively compared the SDNN approach with the state-of-the-art SDP approaches utilizing 10 software defect datasets. The experimental results show that our SDNN is a competitive approach and is able to improve the prediction performance more significantly compared with the benchmarked approaches. INDEX TERMS Siamese dense neural networks, deep learning, metric learning, few-shot learning, software defect prediction.

show abstract

“…In this study, we normalize the data using the most common minimummaximum normalization method. This approach is described in detail [38].…”

Section: ) Data Normalizationmentioning

confidence: 99%

Siamese Dense Neural Network for Software Defect Prediction With Small Data

et al. 2019

View full text Add to dashboard Cite

show abstract

“…By normalizing conventional minimum-redundancy maximum-relevancy (mRMR), Jung et al [23] proposed the NmRMR criterion. They first extracted features from frames to train an initial feature model.…”

Section: Related Workmentioning

confidence: 99%

Noise Invariant Frame Selection: A Simple Method to Address the Background Noise Problem for Text-independent Speaker Verification

Song

Zhang

Schuller

et al. 2018

2018 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

The performance of speaker-related systems usually degrades heavily in practical applications largely due to the presence of background noise. To improve the robustness of such systems in unknown noisy environments, this paper proposes a simple pre-processing method called Noise Invariant Frame Selection (NIFS). Based on several noisy constraints, it selects noise invariant frames from utterances to represent speakers. Experiments conducted on the TIMIT database showed that the NIFS can significantly improve the performance of Vector Quantization (VQ), Gaussian Mixture Model-Universal Background Model (GMM-UBM) and i-vector-based speaker verification systems in different unknown noisy environments with different SNRs, in comparison to their baselines. Meanwhile, the proposed NIFS-based speaker verification systems achieves similar performance when we change the constraints (hyperparameters) or features, which indicates that it is robust and easy to reproduce. Since NIFS is designed as a general algorithm, it could be further applied to other similar tasks.

show abstract