Thailand

In this paper, we present an improved model for voicing silent speech, where audio is synthesized from facial electromyography (EMG) signals. To give our model greater flexibility to learn its own input features, we directly use EMG signals as input in the place of hand-designed features used by prior work. Our model uses convolutional layers to extract features from the signals and Transformer layers to propagate information across longer distances. To provide better signal for learning, we also introduce an auxiliary task of predicting phoneme labels in addition to predicting speech audio features. On an open vocabulary intelligibility evaluation, our model improves the state of the art for this task by an absolute 25.8%.

ACL-IJCNLP 2021

An Improved Model for Voicing Silent Speech

electromyography

silent speech

speech synthesis

speech

**Welcome to ACL-IJCNLP 2021!**

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP). 

As in previous years, the program of the conference includes a poster session, tutorials, workshops and demonstrations in addition to the main conference.


We were able to keep the registration fees similar to those charged for the virtual ACL 2020. The one fee allows attendance at the main conference and any/all tutorials and workshops. These fees would be $125 Regular Early and $175 Regular Late; $50 Student Early and $75 Student Late. Early registration closes at midnight July 11, 2021 (Eastern Daylight Time).

**Reminder:** It is ACL’s policy that at least one author of each accepted paper (including ACL Finding papers) must register for the conference.

**Reminder2:** Underline site will open closer to the event. If you already registered you will receive access detail

Registration is now open

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP).

poster

Sequence-to-sequence models excel at handling natural language variation, but have been shown to struggle with out-of-distribution compositional generalization. This has motivated new specialized architectures with stronger compositional biases, but most of these approaches have only been evaluated on synthetically-generated datasets, which are not representative of natural language variation. In this work we ask: can we develop a semantic parsing approach that handles both natural language variation and compositional generalization? To better assess this capability, we propose new train and test splits of non-synthetic datasets. We demonstrate that strong existing approaches do not perform well across a broad set of evaluations. We also propose NQG-T5, a hybrid model that combines a high-precision grammar-based approach with a pre-trained sequence-to-sequence model. It outperforms existing approaches across several compositional generalization challenges on non-synthetic data, while also being competitive with the state-of-the-art on standard evaluations. While still far from solving this problem, our study highlights the importance of diverse evaluations and the open challenge of handling both compositional generalization and natural language variation in semantic parsing.

Compositional Generalization and Natural Language Variation: Can a Semantic Parsing Approach Handle Both?

We present a targeted, scaled-up comparison of incremental processing in humans and neural language models by collecting by-word reaction time data for sixteen different syntactic test suites across a range of structural phenomena. Human reaction time data comes from a novel online experimental paradigm called the Interpolated Maze task. We compare human reaction times to by-word probabilities for four contemporary language models, with different architectures and trained on a range of data set sizes. We find that across many phenomena, both humans and language models show increased processing difficulty in ungrammatical sentence regions with human and model `accuracy' scores a la Marvin and Linzen (2018) about equal. However, although language model outputs match humans in direction, we show that models systematically under-predict the difference in magnitude of incremental processing difficulty between grammatical and ungrammatical sentences. Specifically, when models encounter syntactic violations they fail to accurately predict the longer reading times observed in the human data. These results call into question whether contemporary language models are approaching human-like performance for sensitivity to syntactic violations.

A Targeted Assessment of Incremental Processing in Neural Language Models and Humans

Part-of-Speech (POS) tags are routinely included as features in many NLP tasks. However, the importance and usefulness of POS tags needs to be examined as NLP expands to low-resource languages because linguists who provide many annotated resources do not place priority on early identification and tagging of POS. This paper describes an empirical study about the effect that POS tags have on two computational morphological tasks with the Transformer architecture. Each task is tested twice on identical data except for the presence/absence of POS tags, using published data in ten high- to low-resource languages or unpublished linguistic field data in five low-resource languages. We find that the presence or absence of POS tags does not have a significant bearing on performance. In joint segmentation and glossing, the largest average difference is an .09 improvement in F1-scores by removing POS tags. In reinflection, the greatest average difference is 1.2% in accuracy for published data and 5% for unpublished and noisy field data.

To POS Tag or Not to POS Tag: The Impact of POS Tags on Morphological Learning in Low-Resource Settings

Parsing spoken dialogue poses unique difficulties, including disfluencies and unmarked boundaries between sentence-like units. Previous work has shown that prosody can help with parsing disfluent speech (Tran et al. 2018), but has assumed that the input to the parser is already segmented into sentence-like units (SUs), which isn't true in existing speech applications. We investigate how prosody affects a parser that receives an entire dialogue turn as input (a turn-based model), instead of gold standard pre-segmented SUs (an SU-based model). In experiments on the English Switchboard corpus, we find that when using transcripts alone, the turn-based model has trouble segmenting SUs, leading to worse parse performance than the SU-based model. However, prosody can effectively replace gold standard SU boundaries: with prosody, the turn-based model performs as well as the SU-based model (91.38 vs. 91.06 F1 score, respectively), despite performing two tasks (SU segmentation and parsing) rather than one (parsing alone). Analysis shows that pitch and intensity features are the most important for this corpus, since they allow the model to correctly distinguish an SU boundary from a speech disfluency -- a distinction that the model otherwise struggles to make.

Prosodic segmentation for parsing spoken dialogue

We introduce VoxPopuli, a large-scale multilingual corpus providing 400K hours of unlabeled speech data in 23 languages. It is the largest open data to date for unsupervised representation learning as well as semi-supervised learning. VoxPopuli also contains 1.8K hours of transcribed speeches in 15 languages and their aligned oral interpretations into 15 target languages totaling 17.3K hours. We provide speech recognition (ASR) baselines and validate the versatility of VoxPopuli unlabeled data in semi-supervised ASR and speech-to-text translation under challenging out-of-domain settings. The corpus is available at https://github.com/facebookresearch/voxpopuli.

VoxPopuli: A Large-Scale Multilingual Speech Corpus for Representation Learning, Semi-Supervised Learning and Interpretation

Link to full paper: https://aclanthology.org/2021.acl-srw.11/

Natural language generation systems have witnessed important progress in the last years, but they are shown to generate tokens that are unrelated to the source input. This problem affects computational models in many NLP tasks, and it is particularly unpleasant in multimodal systems. In this work, we assess the rate of object hallucination in multimodal conversational agents playing the GuessWhat?! referential game. Better visual processing has been shown to mitigate this issue in image captioning; hence, we adapt to the GuessWhat?! task the best visual processing models at disposal, and propose two new models to play the Questioner agent. We show that the new models generate few hallucinations compared to other renowned models available in the literature. Moreover, their hallucinations are less severe (affect task-accuracy less) and are more human-like. We also analyse where hallucinations tend to occur more often through the dialogue: hallucinations are less frequent in earlier turns, cause a cascade hallucination effect, and are often preceded by negative answers, which have been shown to be harder to ground.

“I’ve Seen Things You People Wouldn’t Believe”: Hallucinating Entities in GuessWhat?!

India is one of the most linguistically diverse nations of the world and is culturally very rich. Most of these languages are somewhat similar to each other on account of sharing a common ancestry or being in contact for a long period of time. Nowadays, researchers are constantly putting efforts in utilizing the language relatedness to improve the performance of various NLP systems such as cross lingual semantic search, machine translation, sentiment analysis systems, etc. So in this paper, we performed an extensive case study on similarity involving languages of the Indian subcontinent. Language similarity prediction is defined as the task of measuring how similar the two languages are on the basis of their lexical, morphological and syntactic features. In this study, we concentrate only on the approach to calculate lexical similarity between Indian languages by looking at various factors such as size and type of corpus, similarity algorithms, subword segmentation, etc. The main takeaways from our work are: (i) Relative order of the language similarities largely remain the same, regardless of the factors mentioned above, (ii) Similarity within the same language family is higher, (iii) Languages share more lexical features at the subword level.

How do different factors Impact the Inter-language Similarity? A Case Study on Indian languages ​

Social media is often used by individuals and organisations as a platform to spread misinformation. With the recent coronavirus pandemic we have seen a surge of misinformation on Twitter, posing a danger to public health. In this paper, we compile a large COVID-19misinformation-related Twitter corpus and perform an analysis to discover patterns with respect to vocabulary usage. Among others, our analysis reveals that the variety of topics and vocabulary usage are considerably more limited and negative in tweets related to misinformation than in randomly extracted tweets. In addition to our qualitative analysis, our experimental results show that a simple linear model based only on lexical features is effective in identifying misinformation-related tweets (with accuracy over 80%), providing evidence to the fact that the vocabulary used in misinformation largely differs from generic tweets.

COVID-19 and Misinformation: A Large-Scale Lexical Analysis on Twitter

Currently, text chatting is one of the primary means of communication. However, modern text chat still in general does not offer any navigation or even full-featured search, although the high volumes of messages demand it. In order to mitigate these inconveniences, we formulate the problem of situation-based summarization and propose a special data annotation tool intended for developing training and gold-standard data. 

A situation is a subset of messages revolving around a single event in both temporal and contextual senses: e.g, a group of friends arranging a meeting in chat, agreeing on date, time, and place. Situations can be extracted via information retrieval, natural language processing, and machine learning techniques. Since the task is novel, neither training nor gold-standard datasets for it have been created yet. 

In this paper, we present the formulation of the situation-based summarization problem. Next, we describe Chat Corpora Annotator (CCA): the first annotation system designed specifically for exploring and annotating chat log data. We also introduce a custom query language for semi-automatic situation extraction. Finally, we present the first gold-standard dataset for situation-based summarization. The software source code and the dataset are publicly available

Situation-Based Multiparticipant Chat Summarization: a Concept, an Exploration-Annotation Tool and an Example Collection

In this work, we show that not only the well-known end-of-sequence EOS but also the end-of-paragraph EOP, is part of the semantics of a text, and training with this information improves the text generation itself.

An Improved Model for Voicing Silent Speech

Downloads

Next from ACL-IJCNLP 2021

Compositional Generalization and Natural Language Variation: Can a Semantic Parsing Approach Handle Both?

Similar lecture

Learning Robust Latent Representations for Controllable Speech Synthesis