Abstract

Earthquake-induced landslides can affect people and structures as a result of significant ground shaking or regardless of the intensity. A suitable design ground motion helps to mitigate the impact of such landslides. The most commonly used design ground motion in slope stability analysis is based on probabilistic seismic hazard maps. Uncertainties in the selection of expected ground motion levels have been ignored. The present study is conducted based on improvised fully probabilistic approach, which provides the total probability of slope failure in a particular period under seismic loading by addressing all possible scenarios. This approach is applied to the seismically active Tindharia slope located in Darjeeling, India. The total probability of seismic slope failure obtained in the next 50 years is 30% and the most probable peak-ground acceleration that triggers a landslide is 0.12g. Design peak-ground acceleration predicted from the next 475-year probabilistic seismic hazard map is 1.02g. In the present study, the significant difference in the design peak-ground acceleration from probabilistic seismic-hazard analysis and fully probabilistic approach is observed. The study suggests that the seismic landslide hazard may be overestimated or underestimated when used in the design of ground intensity obtained by conducting PSHA.

Abstract

The northern part of INDIA is most vulnerable in failure of slopes due to hill region with complex geology, high intensity rainfall and frequently occurrence of earthquakes. The stability of slopes has traditionally analyzed using deterministic methods. But uncertainty plays major role in geotechnical engineering the probabilistic characterization of uncertainties in input variables for the assessment of slope stability is very helpful. In this paper, the safety of slope using deterministic and probabilistic methods under static and pseudo static loads is car- ried for Lungchok landslide which is in south district of Sikkim state, India. The Factor of safe- ty calculated by deterministic cannot represent the stability of slope exactly as it is limited to the single value of parameters. The stability of slope is independent of single soil parameter but dependent on random process with distribution of soil parameters. Rather than the conventional factor of safety against the sliding failure the probability of sliding failure is most useful. The stability of slope is analyzed using two approaches deterministic and probabilistic analysis under the effects of static and pseudo static forces. The deterministic approach with Morgen-stern Price limit equilibrium method and the probabilistic approach with Monte Carlo simula- tion have been performed for the stability analysis of Lungcok landslide using Geo-Studio Software. The dependence and efficiency of the probabilistic strategies in the investigation of slope stability was highlighted in this paper.

Abstract

We introduce GEM, a living benchmark for natural language Generation (NLG), its Evaluation, and Metrics. Measuring progress in NLG relies on a constantly evolving ecosystem of automated metrics, datasets, and human evaluation standards. Due to this moving target, new models often still evaluate on divergent anglo-centric corpora with wellestablished, but flawed, metrics. This disconnect makes it challenging to identify the limitations of current models and opportunities for progress. Addressing this limitation, GEM provides an environment in which models can easily be applied to a wide set of tasks and in which evaluation strategies can be tested. Regular updates to the benchmark will help NLG research become more multilingual and evolve the challenge alongside models. This paper serves as the description of the data for which we are organizing a shared task at our ACL 2021 Workshop and to which we invite the entire NLG community to participate.

Abstract

Advances in deep learning have been instrumental in enhancing the performance of face verification systems. Despite their ability to attain high accuracy, most of these systems fail to provide interpretations of their decisions. With the increased demands in making deep learning models more interpretable, numerous post-hoc methods have been proposed to probe the workings of these systems. Yet, the quest for face verification systems that inherently provide interpretations still remains largely unexplored. Additionally, most of the existing face recognition models are highly susceptible to adversarial attacks. In this work, we propose a face verification system which addresses the issue of interpretability by employing modular neural networks. In this, representations for each individual facial parts such as nose, mouth, eyes etc. are learned separately. We also show that our method is significantly more resistant to adversarial attacks, thereby addressing another crucial weakness concerning deep learning models.

Abstract

This paper is a description of the system that was designed by team ”professionals” for the task Offensive Language Identification in Dravidian Languages-EACL 2021. Our system Dravidian Offensive Language Identifier Classifier (DrOLIC) uses Indic-BERT to generate word embeddings which is then fed into a 4- layer Multi Layer Perceptron (MLP) which does the multi-class classification task. The system helped us achieve an F1 score of 85% in the shared task for the Malayalam language.

Abstract

Speech is a natural mode of communication among all beings. India is a densely populated country, and people are diverse throughout the globe. The spoken language is the medium of instruction to interact among the people. The majority of Indian languages are spoken globally. The unavailability of larger volumes of transcribed and annotated speech data is often a hurdle for building reliable speech recognition (ASR) systems for Indian languages. Crowdsourcing strategies are effective in collaboratively collecting speech data resources. This paper describes the experience of large-scale speech data collection for the Telugu language through mobile and web-based applications. With this crowd contributed speech, the performance of the baseline ASR system is shown for clean speech. ASR performance for pink and white noises is also compared for various deep neural network (DNN) based acoustic models. The details regarding the usage of frameworks and their challenges during their implementation are part of this paper. The framework adopted for collecting the speech data is rapid, cost-saving, and offers the advantage of extending it to all the other Indian languages. Index Terms—ASR, Crowd-sourced, TDNN, GMM, SGMM

Abstract

Reinforcement learning (RL) has made significant progress in both abstract and real-world domains, but the majority of state-of-the-art algorithms deal only with monotonic actions. However, some applications require agents to reason over different types of actions. Our application simulates reactionbased molecule generation, used as part of the drug discovery pipeline, and includes both uni-molecular and bi-molecular reactions. This paper introduces a novel framework, towered actor critic (TAC), to handle multiple action types. The TAC framework is general in that it is designed to be combined with any existing RL algorithms for continuous action space. We combine it with TD3 to empirically obtain significantly better results than existing methods in the drug discovery setting. TAC is also applied to RL benchmarks in OpenAI Gym and results show that our framework can improve, or at least does not hurt, performance relative to standard TD3.

Abstract

Object detection, semantic segmentation, and instance seg-mentation form the bases for many computer vision tasks in autonomous driving. The complexity of these tasks increases as we shift from object detection to instance segmentation. The state-of-the-art models are eval-uated on standard datasets such as pascal-voc and ms-cococ, which do not consider the dynamics of road scenes. Driving datasets such as Cityscapes and Berkeley Deep Drive (bdd) are captured in a structured environment with better road markings and fewer variations in the ap-pearance of objects and background. However, the same does not hold for Indian roads. The Indian Driving Dataset (idd) is captured in un-structured driving scenarios and is highly challenging for a model due toits diversity. This work presents a comprehensive evaluation of state-of-the-art models on object detection, semantic segmentation, and instance segmentation on-road scene datasets. We present our analyses and com- pare their quantitative and qualitative performance on structured driv-ing datasets (Cityscapes and bdd) and the unstructured driving dataset (idd); understanding the behavior on these datasets helps in addressing various practical issues and helps in creating real-life applications.

Abstract

Object detection is a key component in autonomous navigation systems that enables localization and classification of the objects in a road scene. Existing object detection methods are trained and inferred on a fixed number of known classes present in road scenes. However, in real-world or open-world road scenes, while inference, we come across unknown objects that the detection model hasn’t seen while training. Hence, we propose O pen Wo rld Object Detection on Road Scenes (ORDER) to address the aforementioned problem for road scenes. Firstly, we introduce Feature-Mix to improve the unknown object detection capabilities of an object detector. Feature-Mix widens the gap between known and unknown classes in latent feature space that helps improve the unknown object detection. Next, we identify that the road scene dataset compared to generic object dataset contains a significant proportion of small objects and has higher intra-class bounding box scale variations, making it challenging to detect the known and unknown objects. We propose a novel loss: Focal regression loss that collectively addresses the problem of small object detection and intra-class bounding box by penalizing more the small bounding boxes and dynamically changing the loss according to object size. Further, the detection of small objects is improved by curriculum learning. Finally, we present an extensive evaluation on two road scene datasets: BDD and IDD. Our experimental evaluations on BDD and IDD shows consistent improvement over the current state-of-the-art method. We believe that this work will lay the foundation for real-world object detection for road scenes.

Abstract

Reference Evapotranspiration (ET0) is one of the prominent hydrologic variables affecting water and energy balances and critical factors for crop water requirements and irrigation scheduling. evapotranspiration is a complex hydrological variable defined by various climatic variables. Various empirical formulations have been developed to estimate ET0 depending upon the availability of meteorological variables. Such empirical formulations are region-specific and are for particular climatic conditions. In this context, mathematical models have emerged as simple and readily implementable for the estimation of ET0 with measured meteorological parameters as independent variables. Such data-driven models can be valuable to predict ET0 when climate data is insufficient. The present study compared various empirical models and data-driven algorithms to predict ET0 using various climate variables. Artificial neural networks (ANN) were adopted to estimate reference ET0. Four empirical methods Penman-Monteith, Hargreaves, Turc, and Priestley-Taylor were used to estimate ET0 at a daily time scale. Dataset consists of daily meteorological data over a period of 51 years (1965–2015) for Hyderabad, the largest city of the Indian state, Telangana, with semi-arid climate. The input variables for the ANN model consist of maximum and minimum air temperatures, relative humidity, solar radiation, and wind speed. The Penman-Monteith method was considered as the standard method to compare the ANN and various empirical models of ET0. ANN model was trained and tested with climate variables as input variables and various empirical models as reference models. The most influencing climate variables on ET0 were found in the order of temperature, solar radiation, wind speed, and relative humidity based on correlation coefficients. These variables have formed as the basis to choose different datasets to train over ANN model. Validation has been carried out using the coefficient of determination (R2) which is obtained for the training (1965–2000) and testing period (2001–2015) period as 0.97 and 0.96 respectively. Temperature and radiation-based models of Turc and Priestley-Taylor methods can be used to estimate ET0 when all other climate variables are not available as they also correlate well with the Penman-Monteith method. Advancement towards artificial intelligence techniques in water resources engineering has motivated to simulate reference ET0 using limited meteorological variables to produce accurate results. Such data-driven algorithms developed based on standard empirical models can be implemented for prediction with limited climate data.