Universal Adversarial Attacks on Spoken Language Assessment Systems

Abstract

There is an increasing demand for automated spoken language assessment (SLA) systems, partly driven by the performance improvements that have come from deep learning based approaches. One aspect of deep learning systems is that they do not require expert derived features, operating directly on the original signal such as a speech recognition (ASR) transcript. This, however, increases their potential susceptibility to adversarial attacks as a form of candidate malpractice. In this paper the sensitivity of SLA systems to a universal black-box attack on the ASR text output is explored. The aim is to obtain a single, universal phrase to maximally increase any candidate's score. Four approaches to detect such adversarial attacks are also described. All the systems, and associated detection approaches, are evaluated on a free (spontaneous) speaking section from a Business English test. It is shown that on deep learning based SLA systems the average candidate score can be increased by almost one grade level using a single six word phrase appended to the end of the response hypothesis. Although these large gains can be obtained, they can be easily detected based on detection shifts from the scores of a "traditional" Gaussian Process based grader.

Minimum Bayes’ Risk Decoding for System Combination of Grammatical Error Correction Systems

“…System Combination for seqseq systems: Individual deep learning systems for classification tasks can be combined in many ways: stacking (Wolpert, 1992), negative correlation learning (Liu and Yao, 1999), max-voter schemes (Ju et al, 2018;Simonyan and Zisserman, 2014) or probability averaging (He et al, 2016;Raina et al, 2020;Szegedy et al, 2015). However, for generative language tasks such as GEC, where the output is a sequence of tokens, many traditional ensembling approaches are inapplicable.…”

Cross-corpora experiments of automatic proficiency assessment and error detection for spoken English

“…Deep learning techniques have brought significant improvements in the field of automatic scoring, for assessing both writing and speaking, such that end-to-end neural based approaches outperformed ETS's SpeechRater (Chen et al, 2018), one of the best known oral proficiency test engines (Xi et al, 2008). Specifically, transformer-based models have led to a remarkable improvement in tasks of predicting linguistic proficiency (Raina et al, 2020;Wang et al, 2021).…”

Proceedings of the 17th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2022)

“…Deep learning techniques have brought significant improvements in the field of automatic scoring, for assessing both writing and speaking, such that end-to-end neural based approaches outperformed ETS's SpeechRater , one of the best known oral proficiency test engines (Xi et al, 2008). Specifically, transformer-based models have led to a remarkable improvement in tasks of predicting linguistic proficiency (Raina et al, 2020;.…”