Mitesh M. Khapra

Sumanth Doddapaneni, Rahul Aralikatte, Gowtham Ramesh et al. · microsoft-research, mila

Building Natural Language Understanding (NLU) capabilities for Indic languages, which have a collective speaker base of more than one billion speakers is absolutely crucial. In this work, we aim to improve the NLU capabilities of Indic languages by making contributions along 3 important axes (i) monolingual corpora (ii) NLU testsets (iii) multilingual LLMs focusing on Indic languages. Specifically, we curate the largest monolingual corpora, IndicCorp, with 20.9B tokens covering 24 languages from 4 language families - a 2.3x increase over prior work, while supporting 12 additional languages. Next, we create a human-supervised benchmark, IndicXTREME, consisting of nine diverse NLU tasks covering 20 languages. Across languages and tasks, IndicXTREME contains a total of 105 evaluation sets, of which 52 are new contributions to the literature. To the best of our knowledge, this is the first effort towards creating a standard benchmark for Indic languages that aims to test the multilingual zero-shot capabilities of pretrained language models. Finally, we train IndicBERT v2, a state-of-the-art model supporting all the languages. Averaged across languages and tasks, the model achieves an absolute improvement of 2 points over a strong baseline. The data and models are available at https://github.com/AI4Bharat/IndicBERT.

Arnav Mhaske, Harshit Kedia, Sumanth Doddapaneni et al. · microsoft-research

We present, Naamapadam, the largest publicly available Named Entity Recognition (NER) dataset for the 11 major Indian languages from two language families. The dataset contains more than 400k sentences annotated with a total of at least 100k entities from three standard entity categories (Person, Location, and, Organization) for 9 out of the 11 languages. The training dataset has been automatically created from the Samanantar parallel corpus by projecting automatically tagged entities from an English sentence to the corresponding Indian language translation. We also create manually annotated testsets for 9 languages. We demonstrate the utility of the obtained dataset on the Naamapadam-test dataset. We also release IndicNER, a multilingual IndicBERT model fine-tuned on Naamapadam training set. IndicNER achieves an F1 score of more than $80$ for $7$ out of $9$ test languages. The dataset and models are available under open-source licences at https://ai4bharat.iitm.ac.in/naamapadam.

Yash Madhani, Sushane Parthan, Priyanka Bedekar et al. · microsoft-research

Transliteration is very important in the Indian language context due to the usage of multiple scripts and the widespread use of romanized inputs. However, few training and evaluation sets are publicly available. We introduce Aksharantar, the largest publicly available transliteration dataset for Indian languages created by mining from monolingual and parallel corpora, as well as collecting data from human annotators. The dataset contains 26 million transliteration pairs for 21 Indic languages from 3 language families using 12 scripts. Aksharantar is 21 times larger than existing datasets and is the first publicly available dataset for 7 languages and 1 language family. We also introduce the Aksharantar testset comprising 103k word pairs spanning 19 languages that enables a fine-grained analysis of transliteration models on native origin words, foreign words, frequent words, and rare words. Using the training set, we trained IndicXlit, a multilingual transliteration model that improves accuracy by 15% on the Dakshina test set, and establishes strong baselines on the Aksharantar testset introduced in this work. The models, mining scripts, transliteration guidelines, and datasets are available at https://github.com/AI4Bharat/IndicXlit under open-source licenses. We hope the availability of these large-scale, open resources will spur innovation for Indic language transliteration and downstream applications. We hope the availability of these large-scale, open resources will spur innovation for Indic language transliteration and downstream applications.

Nandini Mundra, Aditya Nanda Kishore, Raj Dabre et al. · microsoft-research

Language Models (LMs) excel in natural language processing tasks for English but show reduced performance in most other languages. This problem is commonly tackled by continually pre-training and fine-tuning these models for said languages. A significant issue in this process is the limited vocabulary coverage in the original model's tokenizer, leading to inadequate representation of new languages and necessitating an expansion of the tokenizer. The initialization of the embeddings corresponding to new vocabulary items presents a further challenge. Current strategies require cross-lingual embeddings and lack a solid theoretical foundation as well as comparisons with strong baselines. In this paper, we first establish theoretically that initializing within the convex hull of existing embeddings is a good initialization, followed by a novel but simple approach, Constrained Word2Vec (CW2V), which does not require cross-lingual embeddings. Our study evaluates different initialization methods for expanding RoBERTa and LLaMA 2 across four languages and five tasks. The results show that CW2V performs equally well or even better than more advanced techniques. Additionally, simpler approaches like multivariate initialization perform on par with these advanced methods indicating that efficient large-scale multilingual continued pretraining can be achieved even with simpler initialization methods. We release our code publicly (https://github.com/AI4Bharat/VocabAdaptation_LLM/tree/CW2V).

Aman Kumar, Himani Shrotriya, Prachi Sahu et al. · microsoft-research

Natural Language Generation (NLG) for non-English languages is hampered by the scarcity of datasets in these languages. In this paper, we present the IndicNLG Benchmark, a collection of datasets for benchmarking NLG for 11 Indic languages. We focus on five diverse tasks, namely, biography generation using Wikipedia infoboxes, news headline generation, sentence summarization, paraphrase generation and, question generation. We describe the created datasets and use them to benchmark the performance of several monolingual and multilingual baselines that leverage pre-trained sequence-to-sequence models. Our results exhibit the strong performance of multilingual language-specific pre-trained models, and the utility of models trained on our dataset for other related NLG tasks. Our dataset creation methods can be easily applied to modest-resource languages as they involve simple steps such as scraping news articles and Wikipedia infoboxes, light cleaning, and pivoting through machine translation data. To the best of our knowledge, the IndicNLG Benchmark is the first NLG benchmark for Indic languages and the most diverse multilingual NLG dataset, with approximately 8M examples across 5 tasks and 11 languages. The datasets and models are publicly available at https://ai4bharat.iitm.ac.in/indicnlg-suite.

Tahir Javed, Kaushal Santosh Bhogale, Abhigyan Raman et al. · microsoft-research

A cornerstone in AI research has been the creation and adoption of standardized training and test datasets to earmark the progress of state-of-the-art models. A particularly successful example is the GLUE dataset for training and evaluating Natural Language Understanding (NLU) models for English. The large body of research around self-supervised BERT-based language models revolved around performance improvements on NLU tasks in GLUE. To evaluate language models in other languages, several language-specific GLUE datasets were created. The area of speech language understanding (SLU) has followed a similar trajectory. The success of large self-supervised models such as wav2vec2 enable creation of speech models with relatively easy to access unlabelled data. These models can then be evaluated on SLU tasks, such as the SUPERB benchmark. In this work, we extend this to Indic languages by releasing the IndicSUPERB benchmark. Specifically, we make the following three contributions. (i) We collect Kathbath containing 1,684 hours of labelled speech data across 12 Indian languages from 1,218 contributors located in 203 districts in India. (ii) Using Kathbath, we create benchmarks across 6 speech tasks: Automatic Speech Recognition, Speaker Verification, Speaker Identification (mono/multi), Language Identification, Query By Example, and Keyword Spotting for 12 languages. (iii) On the released benchmarks, we train and evaluate different self-supervised models alongside a commonly used baseline FBANK. We show that language-specific fine-tuned models are more accurate than baseline on most of the tasks, including a large gap of 76\% for the Language Identification task. However, for speaker identification, self-supervised models trained on large datasets demonstrate an advantage. We hope IndicSUPERB contributes to the progress of developing speech language understanding models for Indian languages.

Ananya B. Sai, Vignesh Nagarajan, Tanay Dixit et al. · microsoft-research

The rapid growth of machine translation (MT) systems has necessitated comprehensive studies to meta-evaluate evaluation metrics being used, which enables a better selection of metrics that best reflect MT quality. Unfortunately, most of the research focuses on high-resource languages, mainly English, the observations for which may not always apply to other languages. Indian languages, having over a billion speakers, are linguistically different from English, and to date, there has not been a systematic study of evaluating MT systems from English into Indian languages. In this paper, we fill this gap by creating an MQM dataset consisting of 7000 fine-grained annotations, spanning 5 Indian languages and 7 MT systems, and use it to establish correlations between annotator scores and scores obtained using existing automatic metrics. Our results show that pre-trained metrics, such as COMET, have the highest correlations with annotator scores. Additionally, we find that the metrics do not adequately capture fluency-based errors in Indian languages, and there is a need to develop metrics focused on Indian languages. We hope that our dataset and analysis will help promote further research in this area.

Gokul Karthik Kumar, Praveen S, Pratyush Kumar et al.

Deep learning based text-to-speech (TTS) systems have been evolving rapidly with advances in model architectures, training methodologies, and generalization across speakers and languages. However, these advances have not been thoroughly investigated for Indian language speech synthesis. Such investigation is computationally expensive given the number and diversity of Indian languages, relatively lower resource availability, and the diverse set of advances in neural TTS that remain untested. In this paper, we evaluate the choice of acoustic models, vocoders, supplementary loss functions, training schedules, and speaker and language diversity for Dravidian and Indo-Aryan languages. Based on this, we identify monolingual models with FastPitch and HiFi-GAN V1, trained jointly on male and female speakers to perform the best. With this setup, we train and evaluate TTS models for 13 languages and find our models to significantly improve upon existing models in all languages as measured by mean opinion scores. We open-source all models on the Bhashini platform.

Kaushal Santosh Bhogale, Abhigyan Raman, Tahir Javed et al. · microsoft-research

End-to-end (E2E) models have become the default choice for state-of-the-art speech recognition systems. Such models are trained on large amounts of labelled data, which are often not available for low-resource languages. Techniques such as self-supervised learning and transfer learning hold promise, but have not yet been effective in training accurate models. On the other hand, collecting labelled datasets on a diverse set of domains and speakers is very expensive. In this work, we demonstrate an inexpensive and effective alternative to these approaches by ``mining'' text and audio pairs for Indian languages from public sources, specifically from the public archives of All India Radio. As a key component, we adapt the Needleman-Wunsch algorithm to align sentences with corresponding audio segments given a long audio and a PDF of its transcript, while being robust to errors due to OCR, extraneous text, and non-transcribed speech. We thus create Shrutilipi, a dataset which contains over 6,400 hours of labelled audio across 12 Indian languages totalling to 4.95M sentences. On average, Shrutilipi results in a 2.3x increase over publicly available labelled data. We establish the quality of Shrutilipi with 21 human evaluators across the 12 languages. We also establish the diversity of Shrutilipi in terms of represented regions, speakers, and mentioned named entities. Significantly, we show that adding Shrutilipi to the training set of Wav2Vec models leads to an average decrease in WER of 5.8\% for 7 languages on the IndicSUPERB benchmark. For Hindi, which has the most benchmarks (7), the average WER falls from 18.8% to 13.5%. This improvement extends to efficient models: We show a 2.3% drop in WER for a Conformer model (10x smaller than Wav2Vec). Finally, we demonstrate the diversity of Shrutilipi by showing that the model trained with it is more robust to noisy input.

Ashwin Sankar, Srija Anand, Praveen Srinivasa Varadhan et al.

Recent advancements in text-to-speech (TTS) synthesis show that large-scale models trained with extensive web data produce highly natural-sounding output. However, such data is scarce for Indian languages due to the lack of high-quality, manually subtitled data on platforms like LibriVox or YouTube. To address this gap, we enhance existing large-scale ASR datasets containing natural conversations collected in low-quality environments to generate high-quality TTS training data. Our pipeline leverages the cross-lingual generalization of denoising and speech enhancement models trained on English and applied to Indian languages. This results in IndicVoices-R (IV-R), the largest multilingual Indian TTS dataset derived from an ASR dataset, with 1,704 hours of high-quality speech from 10,496 speakers across 22 Indian languages. IV-R matches the quality of gold-standard TTS datasets like LJSpeech, LibriTTS, and IndicTTS. We also introduce the IV-R Benchmark, the first to assess zero-shot, few-shot, and many-shot speaker generalization capabilities of TTS models on Indian voices, ensuring diversity in age, gender, and style. We demonstrate that fine-tuning an English pre-trained model on a combined dataset of high-quality IndicTTS and our IV-R dataset results in better zero-shot speaker generalization compared to fine-tuning on the IndicTTS dataset alone. Further, our evaluation reveals limited zero-shot generalization for Indian voices in TTS models trained on prior datasets, which we improve by fine-tuning the model on our data containing diverse set of speakers across language families. We open-source all data and code, releasing the first TTS model for all 22 official Indian languages.

Akash Kumar Mohankumar, Mitesh M. Khapra

Recent studies have shown the advantages of evaluating NLG systems using pairwise comparisons as opposed to direct assessment. Given $k$ systems, a naive approach for identifying the top-ranked system would be to uniformly obtain pairwise comparisons from all ${k \choose 2}$ pairs of systems. However, this can be very expensive as the number of human annotations required would grow quadratically with $k$. In this work, we introduce Active Evaluation, a framework to efficiently identify the top-ranked system by actively choosing system pairs for comparison using dueling bandit algorithms. We perform extensive experiments with 13 dueling bandits algorithms on 13 NLG evaluation datasets spanning 5 tasks and show that the number of human annotations can be reduced by 80%. To further reduce the number of human annotations, we propose model-based dueling bandit algorithms which combine automatic evaluation metrics with human evaluations. Specifically, we eliminate sub-optimal systems even before the human annotation process and perform human evaluations only on test examples where the automatic metric is highly uncertain. This reduces the number of human annotations required further by 89%. In effect, we show that identifying the top-ranked system requires only a few hundred human annotations, which grow linearly with $k$. Lastly, we provide practical recommendations and best practices to identify the top-ranked system efficiently. Our code has been made publicly available at https://github.com/akashkm99/duelnlg

Tahir Javed, Janki Nawale, Sakshi Joshi et al.

Hindi, one of the most spoken language of India, exhibits a diverse array of accents due to its usage among individuals from diverse linguistic origins. To enable a robust evaluation of Hindi ASR systems on multiple accents, we create a benchmark, LAHAJA, which contains read and extempore speech on a diverse set of topics and use cases, with a total of 12.5 hours of Hindi audio, sourced from 132 speakers spanning 83 districts of India. We evaluate existing open-source and commercial models on LAHAJA and find their performance to be poor. We then train models using different datasets and find that our model trained on multilingual data with good speaker diversity outperforms existing models by a significant margin. We also present a fine-grained analysis which shows that the performance declines for speakers from North-East and South India, especially with content heavy in named entities and specialized terminology.

Kaushal Santosh Bhogale, Deovrat Mehendale, Niharika Parasa et al.

In this study, we tackle the challenge of limited labeled data for low-resource languages in ASR, focusing on Hindi. Specifically, we explore pseudo-labeling, by proposing a generic framework combining multiple ideas from existing works. Our framework integrates multiple base models for transcription and evaluators for assessing audio-transcript pairs, resulting in robust pseudo-labeling for low resource languages. We validate our approach with a new benchmark, IndicYT, comprising diverse YouTube audio files from multiple content categories. Our findings show that augmenting pseudo labeled data from YouTube with existing training data leads to significant performance improvements on IndicYT, without affecting performance on out-of-domain benchmarks, demonstrating the efficacy of pseudo-labeled data in enhancing ASR capabilities for low-resource languages. The benchmark, code and models developed as a part of this work will be made publicly available.

Praveen Srinivasa Varadhan, Ashwin Sankar, Giri Raju et al.

We release Rasa, the first multilingual expressive TTS dataset for any Indian language, which contains 10 hours of neutral speech and 1-3 hours of expressive speech for each of the 6 Ekman emotions covering 3 languages: Assamese, Bengali, & Tamil. Our ablation studies reveal that just 1 hour of neutral and 30 minutes of expressive data can yield a Fair system as indicated by MUSHRA scores. Increasing neutral data to 10 hours, with minimal expressive data, significantly enhances expressiveness. This offers a practical recipe for resource-constrained languages, prioritizing easily obtainable neutral data alongside smaller amounts of expressive data. We show the importance of syllabically balanced data and pooling emotions to enhance expressiveness. We also highlight challenges in generating specific emotions, e.g., fear and surprise.

Emil Biju, Anirudh Sriram, Pratyush Kumar et al.

Self-attention heads are characteristic of Transformer models and have been well studied for interpretability and pruning. In this work, we demonstrate an altogether different utility of attention heads, namely for adversarial detection. Specifically, we propose a method to construct input-specific attention subnetworks (IAS) from which we extract three features to discriminate between authentic and adversarial inputs. The resultant detector significantly improves (by over 7.5%) the state-of-the-art adversarial detection accuracy for the BERT encoder on 10 NLU datasets with 11 different adversarial attack types. We also demonstrate that our method (a) is more accurate for larger models which are likely to have more spurious correlations and thus vulnerable to adversarial attack, and (b) performs well even with modest training sets of adversarial examples.

Emil Biju, Anirudh Sriram, Mitesh M. Khapra et al.

Gesture typing is a method of typing words on a touch-based keyboard by creating a continuous trace passing through the relevant keys. This work is aimed at developing a keyboard that supports gesture typing in Indic languages. We begin by noting that when dealing with Indic languages, one needs to cater to two different sets of users: (i) users who prefer to type in the native Indic script (Devanagari, Bengali, etc.) and (ii) users who prefer to type in the English script but want the output transliterated into the native script. In both cases, we need a model that takes a trace as input and maps it to the intended word. To enable the development of these models, we create and release two datasets. First, we create a dataset containing keyboard traces for 193,658 words from 7 Indic languages. Second, we curate 104,412 English-Indic transliteration pairs from Wikidata across these languages. Using these datasets we build a model that performs path decoding, transliteration, and transliteration correction. Unlike prior approaches, our proposed model does not make co-character independence assumptions during decoding. The overall accuracy of our model across the 7 languages varies from 70-95%.

Srija Anand, Praveen Srinivasa Varadhan, Ashwin Sankar et al.

Publicly available TTS datasets for low-resource languages like Hindi and Tamil typically contain 10-20 hours of data, leading to poor vocabulary coverage. This limitation becomes evident in downstream applications where domain-specific vocabulary coupled with frequent code-mixing with English, results in many OOV words. To highlight this problem, we create a benchmark containing OOV words from several real-world applications. Indeed, state-of-the-art Hindi and Tamil TTS systems perform poorly on this OOV benchmark, as indicated by intelligibility tests. To improve the model's OOV performance, we propose a low-effort and economically viable strategy to obtain more training data. Specifically, we propose using volunteers as opposed to high quality voice artists to record words containing character bigrams unseen in the training data. We show that using such inexpensive data, the model's performance improves on OOV words, while not affecting voice quality and in-domain performance.

Shreya Goyal, Sumanth Doddapaneni, Mitesh M. Khapra et al.

In the past few years, it has become increasingly evident that deep neural networks are not resilient enough to withstand adversarial perturbations in input data, leaving them vulnerable to attack. Various authors have proposed strong adversarial attacks for computer vision and Natural Language Processing (NLP) tasks. As a response, many defense mechanisms have also been proposed to prevent these networks from failing. The significance of defending neural networks against adversarial attacks lies in ensuring that the model's predictions remain unchanged even if the input data is perturbed. Several methods for adversarial defense in NLP have been proposed, catering to different NLP tasks such as text classification, named entity recognition, and natural language inference. Some of these methods not only defend neural networks against adversarial attacks but also act as a regularization mechanism during training, saving the model from overfitting. This survey aims to review the various methods proposed for adversarial defenses in NLP over the past few years by introducing a novel taxonomy. The survey also highlights the fragility of advanced deep neural networks in NLP and the challenges involved in defending them.

Mohammed Safi Ur Rahman Khan, Priyam Mehta, Ananth Sankar et al. · microsoft-research

Despite the considerable advancements in English LLMs, the progress in building comparable models for other languages has been hindered due to the scarcity of tailored resources. Our work aims to bridge this divide by introducing an expansive suite of resources specifically designed for the development of Indic LLMs, covering 22 languages, containing a total of 251B tokens and 74.8M instruction-response pairs. Recognizing the importance of both data quality and quantity, our approach combines highly curated manually verified data, unverified yet valuable data, and synthetic data. We build a clean, open-source pipeline for curating pre-training data from diverse sources, including websites, PDFs, and videos, incorporating best practices for crawling, cleaning, flagging, and deduplication. For instruction-fine tuning, we amalgamate existing Indic datasets, translate/transliterate English datasets into Indian languages, and utilize LLaMa2 and Mixtral models to create conversations grounded in articles from Indian Wikipedia and Wikihow. Additionally, we address toxicity alignment by generating toxic prompts for multiple scenarios and then generate non-toxic responses by feeding these toxic prompts to an aligned LLaMa2 model. We hope that the datasets, tools, and resources released as a part of this work will not only propel the research and development of Indic LLMs but also establish an open-source blueprint for extending such efforts to other languages. The data and other artifacts created as part of this work are released with permissive licenses.

Tahir Javed, Janki Atul Nawale, Eldho Ittan George et al.

We present INDICVOICES, a dataset of natural and spontaneous speech containing a total of 7348 hours of read (9%), extempore (74%) and conversational (17%) audio from 16237 speakers covering 145 Indian districts and 22 languages. Of these 7348 hours, 1639 hours have already been transcribed, with a median of 73 hours per language. Through this paper, we share our journey of capturing the cultural, linguistic and demographic diversity of India to create a one-of-its-kind inclusive and representative dataset. More specifically, we share an open-source blueprint for data collection at scale comprising of standardised protocols, centralised tools, a repository of engaging questions, prompts and conversation scenarios spanning multiple domains and topics of interest, quality control mechanisms, comprehensive transcription guidelines and transcription tools. We hope that this open source blueprint will serve as a comprehensive starter kit for data collection efforts in other multilingual regions of the world. Using INDICVOICES, we build IndicASR, the first ASR model to support all the 22 languages listed in the 8th schedule of the Constitution of India. All the data, tools, guidelines, models and other materials developed as a part of this work will be made publicly available

Ashwin Sankar, Yoach Lacombe, Sherry Thomas et al.

We introduce RASMALAI, a large-scale speech dataset with rich text descriptions, designed to advance controllable and expressive text-to-speech (TTS) synthesis for 23 Indian languages and English. It comprises 13,000 hours of speech and 24 million text-description annotations with fine-grained attributes like speaker identity, accent, emotion, style, and background conditions. Using RASMALAI, we develop IndicParlerTTS, the first open-source, text-description-guided TTS for Indian languages. Systematic evaluation demonstrates its ability to generate high-quality speech for named speakers, reliably follow text descriptions and accurately synthesize specified attributes. Additionally, it effectively transfers expressive characteristics both within and across languages. IndicParlerTTS consistently achieves strong performance across these evaluations, setting a new standard for controllable multilingual expressive speech synthesis in Indian languages.

Ashwin Sankar, Sparsh Jain, Nikhil Narasimhan et al. · microsoft-research

Speech translation for Indian languages remains a challenging task due to the scarcity of large-scale, publicly available datasets that capture the linguistic diversity and domain coverage essential for real-world applications. Existing datasets cover a fraction of Indian languages and lack the breadth needed to train robust models that generalize beyond curated benchmarks. To bridge this gap, we introduce BhasaAnuvaad, the largest speech translation dataset for Indian languages, spanning over 44 thousand hours of audio and 17 million aligned text segments across 14 Indian languages and English. Our dataset is built through a threefold methodology: (a) aggregating high-quality existing sources, (b) large-scale web crawling to ensure linguistic and domain diversity, and (c) creating synthetic data to model real-world speech disfluencies. Leveraging BhasaAnuvaad, we train IndicSeamless, a state-of-the-art speech translation model for Indian languages that performs better than existing models. Our experiments demonstrate improvements in the translation quality, setting a new standard for Indian language speech translation. We will release all the code, data and model weights in the open-source, with permissive licenses to promote accessibility and collaboration.

Janki Atul Nawale, Mohammed Safi Ur Rahman Khan, Janani D et al.

Existing studies on fairness are largely Western-focused, making them inadequate for culturally diverse countries such as India. To address this gap, we introduce INDIC-BIAS, a comprehensive India-centric benchmark designed to evaluate fairness of LLMs across 85 identity groups encompassing diverse castes, religions, regions, and tribes. We first consult domain experts to curate over 1,800 socio-cultural topics spanning behaviors and situations, where biases and stereotypes are likely to emerge. Grounded in these topics, we generate and manually validate 20,000 real-world scenario templates to probe LLMs for fairness. We structure these templates into three evaluation tasks: plausibility, judgment, and generation. Our evaluation of 14 popular LLMs on these tasks reveals strong negative biases against marginalized identities, with models frequently reinforcing common stereotypes. Additionally, we find that models struggle to mitigate bias even when explicitly asked to rationalize their decision. Our evaluation provides evidence of both allocative and representational harms that current LLMs could cause towards Indian identities, calling for a more cautious usage in practical applications. We release INDIC-BIAS as an open-source benchmark to advance research on benchmarking and mitigating biases and stereotypes in the Indian context.

Sumanth Doddapaneni, Mohammed Safi Ur Rahman Khan, Sshubam Verma et al.

Large Language Models (LLMs) are increasingly relied upon to evaluate text outputs of other LLMs, thereby influencing leaderboards and development decisions. However, concerns persist over the accuracy of these assessments and the potential for misleading conclusions. In this work, we investigate the effectiveness of LLMs as evaluators for text generation tasks. We propose FBI, a novel framework designed to examine the proficiency of Evaluator LLMs in assessing four critical abilities in other LLMs: factual accuracy, instruction following, coherence in long-form writing, and reasoning proficiency. By introducing targeted perturbations in answers generated by LLMs, that clearly impact one of these key capabilities, we test whether an Evaluator LLM can detect these quality drops. By creating a total of 2400 perturbed answers covering 22 perturbation categories, we conduct a comprehensive study using different evaluation strategies on five prominent LLMs commonly used as evaluators in the literature. Our findings reveal significant shortcomings in current Evaluator LLMs, which failed to identify quality drops in over 50\% of cases on average. Single-answer and pairwise evaluations demonstrated notable limitations, whereas reference-based evaluations showed comparatively better performance. These results underscore the unreliable nature of current Evaluator LLMs and advocate for cautious implementation in practical applications. Code and data are available at https://github.com/AI4Bharat/FBI.

Jay Gala, Pranjal A. Chitale, Raghavan AK et al.

India has a rich linguistic landscape with languages from 4 major language families spoken by over a billion people. 22 of these languages are listed in the Constitution of India (referred to as scheduled languages) are the focus of this work. Given the linguistic diversity, high-quality and accessible Machine Translation (MT) systems are essential in a country like India. Prior to this work, there was (i) no parallel training data spanning all 22 languages, (ii) no robust benchmarks covering all these languages and containing content relevant to India, and (iii) no existing translation models which support all the 22 scheduled languages of India. In this work, we aim to address this gap by focusing on the missing pieces required for enabling wide, easy, and open access to good machine translation systems for all 22 scheduled Indian languages. We identify four key areas of improvement: curating and creating larger training datasets, creating diverse and high-quality benchmarks, training multilingual models, and releasing models with open access. Our first contribution is the release of the Bharat Parallel Corpus Collection (BPCC), the largest publicly available parallel corpora for Indic languages. BPCC contains a total of 230M bitext pairs, of which a total of 126M were newly added, including 644K manually translated sentence pairs created as part of this work. Our second contribution is the release of the first n-way parallel benchmark covering all 22 Indian languages, featuring diverse domains, Indian-origin content, and source-original test sets. Next, we present IndicTrans2, the first model to support all 22 languages, surpassing existing models on multiple existing and new benchmarks created as a part of this work. Lastly, to promote accessibility and collaboration, we release our models and associated data with permissive licenses at https://github.com/AI4Bharat/IndicTrans2.

Yash Madhani, Mitesh M. Khapra, Anoop Kunchukuttan

We create publicly available language identification (LID) datasets and models in all 22 Indian languages listed in the Indian constitution in both native-script and romanized text. First, we create Bhasha-Abhijnaanam, a language identification test set for native-script as well as romanized text which spans all 22 Indic languages. We also train IndicLID, a language identifier for all the above-mentioned languages in both native and romanized script. For native-script text, it has better language coverage than existing LIDs and is competitive or better than other LIDs. IndicLID is the first LID for romanized text in Indian languages. Two major challenges for romanized text LID are the lack of training data and low-LID performance when languages are similar. We provide simple and effective solutions to these problems. In general, there has been limited work on romanized text in any language, and our findings are relevant to other languages that need romanized language identification. Our models are publicly available at https://ai4bharat.iitm.ac.in/indiclid under open-source licenses. Our training and test sets are also publicly available at https://ai4bharat.iitm.ac.in/bhasha-abhijnaanam under open-source licenses.

Tahir Javed, Sakshi Joshi, Vignesh Nagarajan et al.

India is the second largest English-speaking country in the world with a speaker base of roughly 130 million. Thus, it is imperative that automatic speech recognition (ASR) systems for English should be evaluated on Indian accents. Unfortunately, Indian speakers find a very poor representation in existing English ASR benchmarks such as LibriSpeech, Switchboard, Speech Accent Archive, etc. In this work, we address this gap by creating Svarah, a benchmark that contains 9.6 hours of transcribed English audio from 117 speakers across 65 geographic locations throughout India, resulting in a diverse range of accents. Svarah comprises both read speech and spontaneous conversational data, covering various domains, such as history, culture, tourism, etc., ensuring a diverse vocabulary. We evaluate 6 open source ASR models and 2 commercial ASR systems on Svarah and show that there is clear scope for improvement on Indian accents. Svarah as well as all our code will be publicly available.

Kaushal Santosh Bhogale, Sai Sundaresan, Abhigyan Raman et al.

Improving ASR systems is necessary to make new LLM-based use-cases accessible to people across the globe. In this paper, we focus on Indian languages, and make the case that diverse benchmarks are required to evaluate and improve ASR systems for Indian languages. To address this, we collate Vistaar as a set of 59 benchmarks across various language and domain combinations, on which we evaluate 3 publicly available ASR systems and 2 commercial systems. We also train IndicWhisper models by fine-tuning the Whisper models on publicly available training datasets across 12 Indian languages totalling to 10.7K hours. We show that IndicWhisper significantly improves on considered ASR systems on the Vistaar benchmark. Indeed, IndicWhisper has the lowest WER in 39 out of the 59 benchmarks, with an average reduction of 4.1 WER. We open-source all datasets, code and models.

Nandini Mundra, Sumanth Doddapaneni, Raj Dabre et al.

Adapters have been positioned as a parameter-efficient fine-tuning (PEFT) approach, whereby a minimal number of parameters are added to the model and fine-tuned. However, adapters have not been sufficiently analyzed to understand if PEFT translates to benefits in training/deployment efficiency and maintainability/extensibility. Through extensive experiments on many adapters, tasks, and languages in supervised and cross-lingual zero-shot settings, we clearly show that for Natural Language Understanding (NLU) tasks, the parameter efficiency in adapters does not translate to efficiency gains compared to full fine-tuning of models. More precisely, adapters are relatively expensive to train and have slightly higher deployment latency. Furthermore, the maintainability/extensibility benefits of adapters can be achieved with simpler approaches like multi-task training via full fine-tuning, which also provide relatively faster training times. We, therefore, recommend that for moderately sized models for NLU tasks, practitioners should rely on full fine-tuning or multi-task training rather than using adapters. Our code is available at https://github.com/AI4Bharat/adapter-efficiency.

Prem Selvaraj, Gokul NC, Pratyush Kumar et al.

AI technologies for Natural Languages have made tremendous progress recently. However, commensurate progress has not been made on Sign Languages, in particular, in recognizing signs as individual words or as complete sentences. We introduce OpenHands, a library where we take four key ideas from the NLP community for low-resource languages and apply them to sign languages for word-level recognition. First, we propose using pose extracted through pretrained models as the standard modality of data to reduce training time and enable efficient inference, and we release standardized pose datasets for 6 different sign languages - American, Argentinian, Chinese, Greek, Indian, and Turkish. Second, we train and release checkpoints of 4 pose-based isolated sign language recognition models across all 6 languages, providing baselines and ready checkpoints for deployment. Third, to address the lack of labelled data, we propose self-supervised pretraining on unlabelled data. We curate and release the largest pose-based pretraining dataset on Indian Sign Language (Indian-SL). Fourth, we compare different pretraining strategies and for the first time establish that pretraining is effective for sign language recognition by demonstrating (a) improved fine-tuning performance especially in low-resource settings, and (b) high crosslingual transfer from Indian-SL to few other sign languages. We open-source all models and datasets in OpenHands with a hope that it makes research in sign languages more accessible, available here at https://github.com/AI4Bharat/OpenHands .

Sahana Ramnath, Preksha Nema, Deep Sahni et al.

BERT and its variants have achieved state-of-the-art performance in various NLP tasks. Since then, various works have been proposed to analyze the linguistic information being captured in BERT. However, the current works do not provide an insight into how BERT is able to achieve near human-level performance on the task of Reading Comprehension based Question Answering. In this work, we attempt to interpret BERT for RCQA. Since BERT layers do not have predefined roles, we define a layer's role or functionality using Integrated Gradients. Based on the defined roles, we perform a preliminary analysis across all layers. We observed that the initial layers focus on query-passage interaction, whereas later layers focus more on contextual understanding and enhancing the answer prediction. Specifically for quantifier questions (how much/how many), we notice that BERT focuses on confusing words (i.e., on other numerical quantities in the passage) in the later layers, but still manages to predict the answer correctly. The fine-tuning and analysis scripts will be publicly available at https://github.com/iitmnlp/BERT-Analysis-RCQA .

Anoop Kunchukuttan, Divyanshu Kakwani, Satish Golla et al.

We present the IndicNLP corpus, a large-scale, general-domain corpus containing 2.7 billion words for 10 Indian languages from two language families. We share pre-trained word embeddings trained on these corpora. We create news article category classification datasets for 9 languages to evaluate the embeddings. We show that the IndicNLP embeddings significantly outperform publicly available pre-trained embedding on multiple evaluation tasks. We hope that the availability of the corpus will accelerate Indic NLP research. The resources are available at https://github.com/ai4bharat-indicnlp/indicnlp_corpus.

Akash Kumar Mohankumar, Preksha Nema, Sharan Narasimhan et al.

Recent studies on interpretability of attention distributions have led to notions of faithful and plausible explanations for a model's predictions. Attention distributions can be considered a faithful explanation if a higher attention weight implies a greater impact on the model's prediction. They can be considered a plausible explanation if they provide a human-understandable justification for the model's predictions. In this work, we first explain why current attention mechanisms in LSTM based encoders can neither provide a faithful nor a plausible explanation of the model's predictions. We observe that in LSTM based encoders the hidden representations at different time-steps are very similar to each other (high conicity) and attention weights in these situations do not carry much meaning because even a random permutation of the attention weights does not affect the model's predictions. Based on experiments on a wide variety of tasks and datasets, we observe attention distributions often attribute the model's predictions to unimportant words such as punctuation and fail to offer a plausible explanation for the predictions. To make attention mechanisms more faithful and plausible, we propose a modified LSTM cell with a diversity-driven training objective that ensures that the hidden representations learned at different time steps are diverse. We show that the resulting attention distributions offer more transparency as they (i) provide a more precise importance ranking of the hidden states (ii) are better indicative of words important for the model's predictions (iii) correlate better with gradient-based attribution methods. Human evaluations indicate that the attention distributions learned by our model offer a plausible explanation of the model's predictions. Our code has been made publicly available at https://github.com/akashkm99/Interpretable-Attention

Preksha Nema, Akash Kumar Mohankumar, Mitesh M. Khapra et al.

In this work, we focus on the task of Automatic Question Generation (AQG) where given a passage and an answer the task is to generate the corresponding question. It is desired that the generated question should be (i) grammatically correct (ii) answerable from the passage and (iii) specific to the given answer. An analysis of existing AQG models shows that they produce questions which do not adhere to one or more of {the above-mentioned qualities}. In particular, the generated questions look like an incomplete draft of the desired question with a clear scope for refinement. {To alleviate this shortcoming}, we propose a method which tries to mimic the human process of generating questions by first creating an initial draft and then refining it. More specifically, we propose Refine Network (RefNet) which contains two decoders. The second decoder uses a dual attention network which pays attention to both (i) the original passage and (ii) the question (initial draft) generated by the first decoder. In effect, it refines the question generated by the first decoder, thereby making it more correct and complete. We evaluate RefNet on three datasets, \textit{viz.}, SQuAD, HOTPOT-QA, and DROP, and show that it outperforms existing state-of-the-art methods by 7-16\% on all of these datasets. Lastly, we show that we can improve the quality of the second decoder on specific metrics, such as, fluency and answerability by explicitly rewarding revisions that improve on the corresponding metric during training. The code has been made publicly available \footnote{https://github.com/PrekshaNema25/RefNet-QG}

Preksha Nema, Mitesh M. Khapra

There has always been criticism for using $n$-gram based similarity metrics, such as BLEU, NIST, etc, for evaluating the performance of NLG systems. However, these metrics continue to remain popular and are recently being used for evaluating the performance of systems which automatically generate questions from documents, knowledge graphs, images, etc. Given the rising interest in such automatic question generation (AQG) systems, it is important to objectively examine whether these metrics are suitable for this task. In particular, it is important to verify whether such metrics used for evaluating AQG systems focus on answerability of the generated question by preferring questions which contain all relevant information such as question type (Wh-types), entities, relations, etc. In this work, we show that current automatic evaluation metrics based on $n$-gram similarity do not always correlate well with human judgments about answerability of a question. To alleviate this problem and as a first step towards better evaluation metrics for AQG, we introduce a scoring function to capture answerability and show that when this scoring function is integrated with existing metrics, they correlate significantly better with human judgments. The scripts and data developed as a part of this work are made publicly available at https://github.com/PrekshaNema25/Answerability-Metric

Sumanth Doddapaneni, Mohammed Safi Ur Rahman Khan, Dilip Venkatesh et al. · microsoft-research

Evaluating machine-generated text remains a significant challenge in NLP, especially for non-English languages. Current methodologies, including automated metrics, human assessments, and LLM-based evaluations, predominantly focus on English, revealing a significant gap in multilingual evaluation frameworks. We introduce the Cross Lingual Auto Evaluation (CIA) Suite, an extensible framework that includes evaluator LLMs (Hercule) and a novel test set (Recon) specifically designed for multilingual evaluation. Our test set features 500 human-annotated instructions spanning various task capabilities along with human judgment scores across six languages. This would enable benchmarking of general-purpose multilingual LLMs and facilitate meta-evaluation of Evaluator LLMs. The proposed model, Hercule, is a cross-lingual evaluation model that addresses the scarcity of reference answers in the target language by learning to assign scores to responses based on easily available reference answers in English. Our experiments demonstrate that Hercule aligns more closely with human judgments compared to proprietary models, demonstrating the effectiveness of such cross-lingual evaluation in low resource scenarios. Further, it is also effective in zero-shot evaluation on unseen languages. This study is the first comprehensive examination of cross-lingual evaluation using LLMs, presenting a scalable and effective approach for multilingual assessment. All code, datasets, and models will be publicly available to enable further research in this important area.

Praveen Srinivasa Varadhan, Amogh Gulati, Ashwin Sankar et al.

Despite rapid advancements in TTS models, a consistent and robust human evaluation framework is still lacking. For example, MOS tests fail to differentiate between similar models, and CMOS's pairwise comparisons are time-intensive. The MUSHRA test is a promising alternative for evaluating multiple TTS systems simultaneously, but in this work we show that its reliance on matching human reference speech unduly penalises the scores of modern TTS systems that can exceed human speech quality. More specifically, we conduct a comprehensive assessment of the MUSHRA test, focusing on its sensitivity to factors such as rater variability, listener fatigue, and reference bias. Based on our extensive evaluation involving 492 human listeners across Hindi and Tamil we identify two primary shortcomings: (i) reference-matching bias, where raters are unduly influenced by the human reference, and (ii) judgement ambiguity, arising from a lack of clear fine-grained guidelines. To address these issues, we propose two refined variants of the MUSHRA test. The first variant enables fairer ratings for synthesized samples that surpass human reference quality. The second variant reduces ambiguity, as indicated by the relatively lower variance across raters. By combining these approaches, we achieve both more reliable and more fine-grained assessments. We also release MANGO, a massive dataset of 246,000 human ratings, the first-of-its-kind collection for Indian languages, aiding in analyzing human preferences and developing automatic metrics for evaluating TTS systems.

Srija Anand, Praveen Srinivasa Varadhan, Mehak Singal et al.

Recent advancements in Text-to-Speech (TTS) technology have led to natural-sounding speech for English, primarily due to the availability of large-scale, high-quality web data. However, many other languages lack access to such resources, relying instead on limited studio-quality data. This scarcity results in synthesized speech that often suffers from intelligibility issues, particularly with low-frequency character bigrams. In this paper, we propose three solutions to address this challenge. First, we leverage high-quality data from linguistically or geographically related languages to improve TTS for the target language. Second, we utilize low-quality Automatic Speech Recognition (ASR) data recorded in non-studio environments, which is refined using denoising and speech enhancement models. Third, we apply knowledge distillation from large-scale models using synthetic data to generate more robust outputs. Our experiments with Hindi demonstrate significant reductions in intelligibility issues, as validated by human evaluators. We propose this methodology as a viable alternative for languages with limited access to high-quality data, enabling them to collectively benefit from shared resources.

Tahir Javed, Kaushal Bhogale, Mitesh M. Khapra

We introduce Nirantar, a comprehensive framework for evaluating continual learning (CL) in multilingual and multi-domain ASR. Designed to reflect real-world CL challenges, Nirantar leverages data collected incrementally across 22 languages and 208 districts in India through natural episodes. This enables evaluation across Language-Incremental (LIL), Domain-Incremental (DIL), and the novel Language-Incremental Domain-Incremental Learning (LIDIL) scenarios. Unlike prior work that relies on simulated episodes, Nirantar presents dynamic, non-uniform language and domain shifts, making it an ideal testbed for CL research. With 3250 hours of human-transcribed speech, including 1720 hours newly introduced in this work, our framework enables systematic benchmarking of CL methods. We evaluate existing approaches and demonstrate that no single method performs consistently well, underscoring the need for more robust CL strategies.

Praveen Srinivasa Varadhan, Srija Anand, Soma Siddhartha et al.

What happens when an English Fairytaler is fine-tuned on Indian languages? We evaluate how the English F5-TTS model adapts to 11 Indian languages, measuring polyglot fluency, voice-cloning, style-cloning, and code-mixing. We compare: (i) training from scratch, (ii) fine-tuning English F5 on Indian data, and (iii) fine-tuning on both Indian and English data to prevent forgetting. Fine-tuning with only Indian data proves most effective and the resultant IN-F5 is a near-human polyglot; that enables speakers of one language (e.g., Odia) to fluently speak in another (e.g., Hindi). Our results show English pretraining aids low-resource TTS in reaching human parity. To aid progress in other low-resource languages, we study data-constrained setups and arrive at a compute optimal strategy. Finally, we show IN-F5 can synthesize unseen languages like Bhojpuri and Tulu using a human-in-the-loop approach for zero-resource TTS via synthetic data generation.

Anushka Singh, Ananya B. Sai, Raj Dabre et al.

While machine translation evaluation has been studied primarily for high-resource languages, there has been a recent interest in evaluation for low-resource languages due to the increasing availability of data and models. In this paper, we focus on a zero-shot evaluation setting focusing on low-resource Indian languages, namely Assamese, Kannada, Maithili, and Punjabi. We collect sufficient Multi-Dimensional Quality Metrics (MQM) and Direct Assessment (DA) annotations to create test sets and meta-evaluate a plethora of automatic evaluation metrics. We observe that even for learned metrics, which are known to exhibit zero-shot performance, the Kendall Tau and Pearson correlations with human annotations are only as high as 0.32 and 0.45. Synthetic data approaches show mixed results and overall do not help close the gap by much for these languages. This indicates that there is still a long way to go for low-resource evaluation.

Jay Gala, Thanmay Jayakumar, Jaavid Aktar Husain et al.

We announce the initial release of "Airavata," an instruction-tuned LLM for Hindi. Airavata was created by fine-tuning OpenHathi with diverse, instruction-tuning Hindi datasets to make it better suited for assistive tasks. Along with the model, we also share the IndicInstruct dataset, which is a collection of diverse instruction-tuning datasets to enable further research for Indic LLMs. Additionally, we present evaluation benchmarks and a framework for assessing LLM performance across tasks in Hindi. Currently, Airavata supports Hindi, but we plan to expand this to all 22 scheduled Indic languages. You can access all artifacts at https://ai4bharat.github.io/airavata.

Tahir Javed, Sumanth Doddapaneni, Abhigyan Raman et al.

Recent methods in speech and language technology pretrain very LARGE models which are fine-tuned for specific tasks. However, the benefits of such LARGE models are often limited to a few resource rich languages of the world. In this work, we make multiple contributions towards building ASR systems for low resource languages from the Indian subcontinent. First, we curate 17,000 hours of raw speech data for 40 Indian languages from a wide variety of domains including education, news, technology, and finance. Second, using this raw speech data we pretrain several variants of wav2vec style models for 40 Indian languages. Third, we analyze the pretrained models to find key features: codebook vectors of similar sounding phonemes are shared across languages, representations across layers are discriminative of the language family, and attention heads often pay attention within small local windows. Fourth, we fine-tune this model for downstream ASR for 9 languages and obtain state-of-the-art results on 3 public datasets, including on very low-resource languages such as Sinhala and Nepali. Our work establishes that multilingual pretraining is an effective strategy for building ASR systems for the linguistically diverse speakers of the Indian subcontinent. Our code, data and models are available publicly at https://indicnlp.ai4bharat.org/indicwav2vec/ and we hope they will help advance research in ASR for Indic languages.

Sahana Ramnath, Preksha Nema, Deep Sahni et al.

As neural-network-based QA models become deeper and more complex, there is a demand for robust frameworks which can access a model's rationale for its prediction. Current techniques that provide insights on a model's working are either dependent on adversarial datasets or are proposing models with explicit explanation generation components. These techniques are time-consuming and challenging to extend to existing models and new datasets. In this work, we use `Integrated Gradients' to extract rationale for existing state-of-the-art models in the task of Reading Comprehension based Question Answering (RCQA). On detailed analysis and comparison with collected human rationales, we find that though ~40-80% words of extracted rationale coincide with the human rationale (precision), only 6-19% of human rationale is present in the extracted rationale (recall).

Aakriti Budhraja, Madhura Pande, Pratyush Kumar et al.

Large multilingual models, such as mBERT, have shown promise in crosslingual transfer. In this work, we employ pruning to quantify the robustness and interpret layer-wise importance of mBERT. On four GLUE tasks, the relative drops in accuracy due to pruning have almost identical results on mBERT and BERT suggesting that the reduced attention capacity of the multilingual models does not affect robustness to pruning. For the crosslingual task XNLI, we report higher drops in accuracy with pruning indicating lower robustness in crosslingual transfer. Also, the importance of the encoder layers sensitively depends on the language family and the pre-training corpus size. The top layers, which are relatively more influenced by fine-tuning, encode important information for languages similar to English (SVO) while the bottom layers, which are relatively less influenced by fine-tuning, are particularly important for agglutinative and low-resource languages.

Ananya B. Sai, Tanay Dixit, Dev Yashpal Sheth et al.

Natural Language Generation (NLG) evaluation is a multifaceted task requiring assessment of multiple desirable criteria, e.g., fluency, coherency, coverage, relevance, adequacy, overall quality, etc. Across existing datasets for 6 NLG tasks, we observe that the human evaluation scores on these multiple criteria are often not correlated. For example, there is a very low correlation between human scores on fluency and data coverage for the task of structured data to text generation. This suggests that the current recipe of proposing new automatic evaluation metrics for NLG by showing that they correlate well with scores assigned by humans for a single criteria (overall quality) alone is inadequate. Indeed, our extensive study involving 25 automatic evaluation metrics across 6 different tasks and 18 different evaluation criteria shows that there is no single metric which correlates well with human scores on all desirable criteria, for most NLG tasks. Given this situation, we propose CheckLists for better design and evaluation of automatic metrics. We design templates which target a specific criteria (e.g., coverage) and perturb the output such that the quality gets affected only along this specific criteria (e.g., the coverage drops). We show that existing evaluation metrics are not robust against even such simple perturbations and disagree with scores assigned by humans to the perturbed output. The proposed templates thus allow for a fine-grained assessment of automatic evaluation metrics exposing their limitations and will facilitate better design, analysis and evaluation of such metrics.

Raj Dabre, Himani Shrotriya, Anoop Kunchukuttan et al.

In this paper, we study pre-trained sequence-to-sequence models for a group of related languages, with a focus on Indic languages. We present IndicBART, a multilingual, sequence-to-sequence pre-trained model focusing on 11 Indic languages and English. IndicBART utilizes the orthographic similarity between Indic scripts to improve transfer learning between similar Indic languages. We evaluate IndicBART on two NLG tasks: Neural Machine Translation (NMT) and extreme summarization. Our experiments on NMT and extreme summarization show that a model specific to related languages like IndicBART is competitive with large pre-trained models like mBART50 despite being significantly smaller. It also performs well on very low-resource translation scenarios where languages are not included in pre-training or fine-tuning. Script sharing, multilingual training, and better utilization of limited model capacity contribute to the good performance of the compact IndicBART model.

Sumanth Doddapaneni, Gowtham Ramesh, Mitesh M. Khapra et al.

Multilingual Language Models (\MLLMs) such as mBERT, XLM, XLM-R, \textit{etc.} have emerged as a viable option for bringing the power of pretraining to a large number of languages. Given their success in zero-shot transfer learning, there has emerged a large body of work in (i) building bigger \MLLMs~covering a large number of languages (ii) creating exhaustive benchmarks covering a wider variety of tasks and languages for evaluating \MLLMs~ (iii) analysing the performance of \MLLMs~on monolingual, zero-shot cross-lingual and bilingual tasks (iv) understanding the universal language patterns (if any) learnt by \MLLMs~ and (v) augmenting the (often) limited capacity of \MLLMs~ to improve their performance on seen or even unseen languages. In this survey, we review the existing literature covering the above broad areas of research pertaining to \MLLMs. Based on our survey, we recommend some promising directions of future research.

Madhura Pande, Aakriti Budhraja, Preksha Nema et al.

Multi-headed attention heads are a mainstay in transformer-based models. Different methods have been proposed to classify the role of each attention head based on the relations between tokens which have high pair-wise attention. These roles include syntactic (tokens with some syntactic relation), local (nearby tokens), block (tokens in the same sentence) and delimiter (the special [CLS], [SEP] tokens). There are two main challenges with existing methods for classification: (a) there are no standard scores across studies or across functional roles, and (b) these scores are often average quantities measured across sentences without capturing statistical significance. In this work, we formalize a simple yet effective score that generalizes to all the roles of attention heads and employs hypothesis testing on this score for robust inference. This provides us the right lens to systematically analyze attention heads and confidently comment on many commonly posed questions on analyzing the BERT model. In particular, we comment on the co-location of multiple functional roles in the same attention head, the distribution of attention heads across layers, and effect of fine-tuning for specific NLP tasks on these functional roles.

Dev Yashpal Sheth, Sreyas Mohan, Joshua L. Vincent et al.

Deep convolutional neural networks (CNNs) for video denoising are typically trained with supervision, assuming the availability of clean videos. However, in many applications, such as microscopy, noiseless videos are not available. To address this, we propose an Unsupervised Deep Video Denoiser (UDVD), a CNN architecture designed to be trained exclusively with noisy data. The performance of UDVD is comparable to the supervised state-of-the-art, even when trained only on a single short noisy video. We demonstrate the promise of our approach in real-world imaging applications by denoising raw video, fluorescence-microscopy and electron-microscopy data. In contrast to many current approaches to video denoising, UDVD does not require explicit motion compensation. This is advantageous because motion compensation is computationally expensive, and can be unreliable when the input data are noisy. A gradient-based analysis reveals that UDVD automatically adapts to local motion in the input noisy videos. Thus, the network learns to perform implicit motion compensation, even though it is only trained for denoising.