Abstract

This study presents an Internet of Things (IoT)- based system that utilises machine learning (ML) techniques to estimate water flow through pipes based on pressure. The system incorporates an ESP-32 microcontroller, a Danfoss MBS 3000 pressure sensor, and a flow meter deployed at three locations to collect data for three months. To model the relationship between pressure and flow rate, ML algorithms such as linear regression (LR), support vector regression (SVR), and convolutional neural network (CNN) were trained, analysed, and compared. By establishing a model to estimate the flow rate based on pressure, the need for a flow meter in the setup can be eliminated. The system’s low-cost, easy-to-implement, and non-invasive nature makes it suitable for widespread adoption in residential areas offering a promising solution for optimising water distribution and reducing water wastage.

Abstract

The high rate of groundwater usage has resulted in a rapid decline in groundwater levels, which has necessitated its monitoring. This paper focuses on estimating the groundwater level inside borewells remotely and in real-time using the Internet of Things (IoT). For this, a solution is proposed which is based on string tension and does not require any electrical components to be lowered into the borewell. The solution provides a completely automated, real-time, reliable, and IoT-enabled alternative to existing methods for borewell water monitoring. The proposed approach is compared with an ultrasonic sensor-based approach in a controlled environment as well as in a tank. Additionally, four nodes were deployed in a small educational campus in the Indian city of Hyderabad to ascertain its usability and reliability in practical situations. The observations from the data collected over a month show that the proposed low-cost solution is reliable and has a good performance in the field. Index Terms—Borewell, Real-time, IoT, String tension, Water-level.

Abstract

This paper addresses the challenge of active speaker detection in automated video editing and highlights the limitations of current audio-only and audio-visual speaker detection methods in handling unseen data with overlapped speakers, speaker occlusions, low video resolution, and random noises. Firstly, we select the BBC Old School Dataset, a comprehensive dataset introduced for automated video editing, and annotate it with active speaker labels. We propose an audio-based nearest neigh- bour algorithm that utilizes additional inputs, such as audio samples of each speaker and faces, to predict and track the active speaker. We evaluate the effectiveness of our approach on the BBC Old School Dataset by utilizing f rame-level speaker accu- racy, which we consider a more suitable metric in the context of video editing. We observe that this simple setup outperforms the current state-of-the-art methods in predicting the active speaker. By incorporating these methods into our speaker-based editing approaches, we also notice that our method closely approximates the output obtained using ground truth annotations

Abstract

Eliminating time-consuming post-production processes and delivering high-quality videos in today’s fast-paced digital landscape are the key advantages of real-time ap- proaches. To address these needs, we present Real Time GAZED: a real-time adaptation of the GAZED framework integrated with CineFilter, a novel real-time camera trajec- tory stabilization approach. It enables users to create pro- fessionally edited videos in real-time. Comparative evalua- tions against baseline methods, including the non-real-time GAZED, demonstrate that Real Time GAZED achieves simi- lar editing results, ensuring high-quality video output. Fur- thermore, a user study confirms the aesthetic quality of the video edits produced by the Real Time GAZED approach. With these advancements in real-time camera trajectory op- timization and video editing presented, the demand for im- mediate and dynamic content creation in industries such as live broadcasting, sports coverage, news reporting, and so- cial media content creation can be met more efficiently

Abstract

Unmanned Aerial Vehicles (UAVs), particularly low-cost UAVs, have become increasingly important due to their wide range of applications and ease of use. However, with the rapid growth of the UAV market, the rising security concerns pose a greater risk. One such primary concern is location spoofing attacks which can compromise UAV's navigation system, making it crucial to analyze various location-based attacks. In this paper, we identify 16 such GPS spoofing scenarios based on environmental conditions, attack type, and spoof signal propagation path. We evaluate these scenarios based on various GPS parameters like Horizontal Dilution Of Precision (HDOP), Vertical Dilution Of Precision (VDOP), GPS satellite count in view, and avg signal-to-noise power density (CN0). We then analyze the variations in GPS parameters for various such attack scenarios. Further, we analyze the impact of distance on average CN0 and the effect of satellite count on effective spoofable distance. We also discuss several critical insights which are empirically observed during our experimental trials. Our experiments revealed that the natural conditions within indoor and outdoor scenarios can vary considerably, and effective spoofable distance can be up to 100 meters when the satellite count is less than 10.

Abstract

With increasing connectivity and sophisticated software, modern vehicles are able to leverage different kinds of services provided by the environment. One such service recommended by the Automotive Edge Computing Consortium (AECC) is downloading high-definition map data by vehicles. This high volume of data can be provided to the vehicles when moving by pre-allocating resources on edge server nodes or roadside units if the routes are known apriori. However, this is not a realistic assumption to make in general. Therefore, in this work, we propose a two-stage optimization framework for efficient data delivery to connected vehicles via edge nodes while considering dynamic route changes. We have evaluated the efficiency of this proposed approach (considering a real-world dataset) with respect to (a) offline optimization strategies considering fixed routes and (b) a greedy approach considering route changes. Our proposed approach works considerably better than the existing approaches in the context of dynamic route changes.

Abstract

The evacuation path from inside a building to safe point outside becomes highly unpredictable due to changes in the local geometry or presence of obstacles during a disaster like a fire. During emergencies, Evacuees need appropriate information and hence prediction of an unobstructed path, as it emerges, needs to be computed well. Understanding the exits with its allowable people flow rate; the type - door or alternative exit such as windows, balconies, etc.; and its role as a node in the graph network is important to ensure safe and timely evacuation from a building. The study here evaluates how obstacles present in the evacuation route affect the removal of the last person. These obstacles, such as furniture, decrease the flow rate at which evacuees can escape. A subspace model is proposed for geometric spaces or carpet areas containing obstacles and is used to compute the shortest obstacle-free paths. The occupancy is considered within the subspaces containing obstacles. The proposed method clearly shows that a graphbased path generation using a subspace model improves the computation time, can be dynamically adapted, and can be scalable across geometric spaces. The results clearly show the impact of the obstacles, with a 2x to 6x rise when compared to obstacle-free scenarios.

Abstract

Social media provides a supportive and anonymous environment for discussing mental health issues, including depression. Existing research on identifying the cause of depression focuses primarily on improving classifier models, while neglecting the importance of learning better data representations. To address this gap, we introduce an architecture that enhances the identification of the cause of depression by learning improved data representations. Our work enables a deeper interpretation of the cause of depression in social media contexts, emphasizing the significance of effective rep- resentation learning for this task. Our work can act as a foundation for self-help applications in the field of mental health.

Abstract

This paper introduces a novel method for determining the best room to place an object in, for embodied scene rearrangement. While state-of-the-art approaches rely on large language models (LLMs) or reinforcement learned (RL) policies for this task, our approach, CLIPGraphs, efficiently combines commonsense domain knowledge, data-driven methods, and recent advances in multimodal learning. Specifically, it (a) encodes a knowledge graph of prior human preferences about the room location of different objects in home environments, (b) incorporates vision-language features to support multimodal queries based on images or text, and (c) uses a graph network to learn object-room affinities based on embeddings of the prior knowledge and the vision-language features. We demonstrate that our approach provides better estimates of the most appropriate location of objects from a benchmark set of object categories in comparison with state-of-the-art baselines

Abstract

Significant progress has been made in speaker-dependent Lip-to-Speech synthesis, which aims to generate speech from silent videos of talking faces. Current state-of-the-art approaches primarily employ non-autoregressive sequence-to-sequence architectures to directly predict mel-spectrograms or audio waveforms from lip representations. We hypothesize that the direct mel-prediction hampers training/model efficiency due to the entanglement of speech content with ambient information and speaker characteristics. To this end, we propose RobustL2S, a modularized framework for Lip-to-Speech synthesis. First, a non-autoregressive sequence-to-sequence model maps self-supervised visual features to a representation of disentangled speech content. A vocoder then converts the speech features into raw waveforms. Extensive evaluations confirm the effectiveness of our setup, achieving state-of-the-art performance on the unconstrained Lip2Wav dataset and the constrained GRID and TCD-TIMIT datasets. Speech samples from RobustL2S can be found at https://neha-sherin.github.io/RobustL2S/