A comparative study of robustness of deep learning approaches for VAD

Tong, Sibo; Gu, Hao; Yu, Kai

doi:10.1109/icassp.2016.7472768

Cited by 41 publications

(13 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This makes it hard for an attacker to change an output decision without drastically changing the input (which may significantly hurt the semantic information and thus is infeasible for the attacker). RNN detectors are temporally deeper and can model long-range temporal features, thus, are more robust to adversarial attacks [37].…”

Section: B Adversarial Robustness Evaluation 1) Impact Of Dnn Architecturementioning

confidence: 99%

All Your Fake Detector are Belong to Us: Evaluating Adversarial Robustness of Fake-News Detectors Under Black-Box Settings

et al. 2021

View full text Add to dashboard Cite

With the hyperconnectivity and ubiquity of the Internet, the fake news problem now presents a greater threat than ever before. One promising solution for countering this threat is to leverage deep learning (DL)-based text classification methods for fake-news detection. However, since such methods have been shown to be vulnerable to adversarial attacks, the integrity and security of DL-based fake news classifiers are under question. Although many works study text classification under the adversarial threat, to the best of our knowledge, we do not find any work in literature that specifically analyzes the performance of DL-based fake-news detectors under adversarial settings. We bridge this gap by evaluating the performance of fake-news detectors under various configurations under black-box settings. In particular, we investigate the robustness of four different DL architectural choices-multilayer perceptron (MLP), convolutional neural network (CNN), recurrent neural network (RNN) and a recently proposed Hybrid CNN-RNN trained on three different state-of-the-art datasets-under different adversarial attacks (Text Bugger, Text Fooler, PWWS, and Deep Word Bug) implemented using the state-of-the-art NLP attack library, Text-Attack. Additionally, we explore how changing the detector complexity, the input sequence length, and the training loss affect the robustness of the learned model. Our experiments suggest that RNNs are robust as compared to other architectures. Further, we show that increasing the input sequence length generally increases the detector's robustness. Our evaluations provide key insights to robustify fake-news detectors against adversarial attacks.INDEX TERMS fake news detection, deep neural networks, adversarial attacks, adversarial robustness.

show abstract

Section: B Adversarial Robustness Evaluation 1) Impact Of Dnn Architecturementioning

confidence: 99%

All Your Fake Detector are Belong to Us: Evaluating Adversarial Robustness of Fake-News Detectors Under Black-Box Settings

et al. 2021

View full text Add to dashboard Cite

show abstract

Section: B Adversarial Robustness Evaluationmentioning

confidence: 99%

All Your Fake Detector Are Belong to Us: Evaluating Adversarial Robustness of Fake-news Detectors Under Black-Box Settings

Argani¹,

Khan²,

AlGhadhban³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…F (•) ∈ R 2 denotes the neural network based function that outputs the conditional probability of each state, θ denotes its parameter, and [•]s t is the stth element of the vector. A long shortterm memory (LSTM) network is often used as F (•) to handle long-term dependencies between input features [27]. Cross entropy (CE) is often used as a loss function to estimate the parameter θ that maximizes (1) as…”

Section: Standard Voice Activity Detectionmentioning

confidence: 99%

Enrollment-Less Training for Personalized Voice Activity Detection

Makishima¹,

Ihori²,

Tanaka³

et al. 2021

Interspeech 2021

View full text Add to dashboard Cite

We present a novel personalized voice activity detection (PVAD) learning method that does not require enrollment data during training. PVAD is a task to detect the speech segments of a specific target speaker at the frame level using enrollment speech of the target speaker. Since PVAD must learn speakers' speech variations to clarify the boundary between speakers, studies on PVAD used large-scale datasets that contain many utterances for each speaker. However, the datasets to train a PVAD model are often limited because substantial cost is needed to prepare such a dataset. In addition, we cannot utilize the datasets used to train the standard VAD because they often lack speaker labels. To solve these problems, our key idea is to use one utterance as both a kind of enrollment speech and an input to the PVAD during training, which enables PVAD training without enrollment speech. In our proposed method, called enrollment-less training, we augment one utterance so as to create variability between the input and the enrollment speech while keeping the speaker identity, which avoids the mismatch between training and inference. Our experimental results demonstrate the efficacy of the method.

show abstract

A comparative study of robustness of deep learning approaches for VAD

Cited by 41 publications

References 18 publications

All Your Fake Detector are Belong to Us: Evaluating Adversarial Robustness of Fake-News Detectors Under Black-Box Settings

All Your Fake Detector are Belong to Us: Evaluating Adversarial Robustness of Fake-News Detectors Under Black-Box Settings

All Your Fake Detector Are Belong to Us: Evaluating Adversarial Robustness of Fake-news Detectors Under Black-Box Settings

Enrollment-Less Training for Personalized Voice Activity Detection

Contact Info

Product

Resources

About