Toxic speech, also known as hate speech, is regarded as one of the crucial issues plaguing online social media today. Most recent work on toxic speech detection is constrained to the modality of text and written conversations with very limited work on toxicity detection from spoken utterances or using the modality of speech. In this paper, we introduce a new dataset DeToxy, the first publicly available toxicity annotated dataset for the English language. DeToxy is sourced from various openly available speech databases and consists of over 2 million utterances. We believe that our dataset would act as a benchmark for the relatively new and un-explored Spoken Language Processing task of detecting toxicity from spoken utterances and boost further research in this space. Finally, we also provide strong unimodal baselines for our dataset and compare traditional two-step and E2E approaches. Our experiments show that in the case of spoken utterances, text-based approaches are largely dependent on gold human-annotated transcripts for their performance and also suffer from the problem of keyword bias. However, the presence of speech files in DeToxy helps facilitates the development of E2E speech models which alleviate both the above-stated problems by better capturing speech clues.

Expression of emotions is a crucial part of daily human communication. Emotion recognition in conversations (ERC) is an emerging field of study, where the primary task is to identify the emotion behind each utterance in a conversation. Though a lot of work has been done on ERC in the past, these works only focus on ERC in the English language, thereby ignoring any other languages. In this paper, we present Multilingual MELD (M-MELD), where we extend the Multimodal EmotionLines Dataset (MELD) \cite{poria2018meld} to 4 other languages beyond English, namely Greek, Polish, French, and Spanish. Beyond just establishing strong baselines for all of these 4 languages, we also propose a novel architecture, DiscLSTM, that uses both sequential and conversational discourse context in a conversational dialogue for ERC. Our proposed approach is computationally efficient, can transfer across languages using just a cross-lingual encoder, and achieves better performance than most uni-modal text approaches in the literature on both MELD and M-MELD. We make our data and code publicly on GitHub.

The users of the Internet increase every moment with increasing population and accessibility of the Internet. With the increase in the number of users of the Internet, the number of controversies, arguments and abuses of all kinds increases. It becomes necessary for social media and other sites to identify toxic content amongst a large number of content being posted by the users of the sites every second. The traditional algorithms that depend on users reporting toxic content for it to be deleted and necessary actions to be taken against the users posting the content would take a long time, within which it would have gained media attention and would have lead to huge fights over the content. Thus, it becomes important for the content to be evaluated for toxicity at the time it is posted in order to stop it from being posted. Therefore, we have designed and trained a deep learning model that can be read through the textual content given through it and determine if it is toxic or not.

With increasing population levels and poverty rate it has become a major problem for Non-profit organizations and agencies to ensure that the right kind of people receive alleviation. The world's poorest typically cannot provide the necessary income and expense records to prove that they qualify for aid. In Latin America, one popular method to verify income qualification is by using an algorithm called the Proxy Means Test (or PMT). With PMT, agencies use a model that considers a family's observable household attributes like the material of their walls and ceiling, or the assets found in the home to classify them and predict their level of need. In this paper we aim to examine and analyse how the individual and household characteristics attribute to poverty levels in Costa Rica and ensemble the results of various machine learning models to achieve better accuracy.

Every year, an increasingly large number of neonatal deaths occur in India. Premature birth and asphyxia are being two of the leading causes of these neonatal deaths. A well-regulated thermal environment is critical for neonatal survival. In the current scenario, it is impossible for the health centers in the rural areas of India to afford a neonatal incubator for every newborn due to its price and transportability. The successful delivery of neonates is hampered in India due to its increasing population along with limited technology and resources. Thus, a prototype of an incubator has been designed that is affordable, transportable, and energy saving for the health centers in the rural regions, with an AI-based decision support system.