-
HOPS: High-order Polynomials with Self-supervised Dimension Reduction for Load Forecasting
Authors:
Pengyang Song,
Han Feng,
Shreyashi Shukla,
Jue Wang,
Tao Hong
Abstract:
Load forecasting is a fundamental task in smart grid. Many techniques have been applied to developing load forecasting models. Due to the challenges such as the Curse of Dimensionality, overfitting, and limited computing resources, multivariate higher-order polynomial models have received limited attention in load forecasting, despite their desirable mathematical foundations and optimization prope…
▽ More
Load forecasting is a fundamental task in smart grid. Many techniques have been applied to developing load forecasting models. Due to the challenges such as the Curse of Dimensionality, overfitting, and limited computing resources, multivariate higher-order polynomial models have received limited attention in load forecasting, despite their desirable mathematical foundations and optimization properties. In this paper, we propose low rank approximation and self-supervised dimension reduction to address the aforementioned issues. To further improve computational efficiency, we also utilize a fast Conjugate Gradient based algorithm for the proposed polynomial models. Based on the load datasets from the ISO New England, the proposed method high-order polynomials with self-supervised dimension reduction (HOPS) demonstrates higher forecasting accuracy over several competitive models. Additionally, experimental results indicate that our approach alleviates redundant variable construction, achieving better forecasts with fewer input variables.
△ Less
Submitted 12 March, 2025; v1 submitted 17 January, 2025;
originally announced January 2025.
-
Advancing Topic Segmentation of Broadcasted Speech with Multilingual Semantic Embeddings
Authors:
Sakshi Deo Shukla,
Pavel Denisov,
Tugtekin Turan
Abstract:
Recent advancements in speech-based topic segmentation have highlighted the potential of pretrained speech encoders to capture semantic representations directly from speech. Traditionally, topic segmentation has relied on a pipeline approach in which transcripts of the automatic speech recognition systems are generated, followed by text-based segmentation algorithms. In this paper, we introduce an…
▽ More
Recent advancements in speech-based topic segmentation have highlighted the potential of pretrained speech encoders to capture semantic representations directly from speech. Traditionally, topic segmentation has relied on a pipeline approach in which transcripts of the automatic speech recognition systems are generated, followed by text-based segmentation algorithms. In this paper, we introduce an end-to-end scheme that bypasses this conventional two-step process by directly employing semantic speech encoders for segmentation. Focused on the broadcasted news domain, which poses unique challenges due to the diversity of speakers and topics within single recordings, we address the challenge of accessing topic change points efficiently in an end-to-end manner. Furthermore, we propose a new benchmark for spoken news topic segmentation by utilizing a dataset featuring approximately 1000 hours of publicly available recordings across six European languages and including an evaluation set in Hindi to test the model's cross-domain performance in a cross-lingual, zero-shot scenario. This setup reflects real-world diversity and the need for models adapting to various linguistic settings. Our results demonstrate that while the traditional pipeline approach achieves a state-of-the-art $P_k$ score of 0.2431 for English, our end-to-end model delivers a competitive $P_k$ score of 0.2564. When trained multilingually, these scores further improve to 0.1988 and 0.2370, respectively. To support further research, we release our model along with data preparation scripts, facilitating open research on multilingual spoken news topic segmentation.
△ Less
Submitted 10 September, 2024;
originally announced September 2024.
-
Performance of Integrated IoT Network with Hybrid mmWave/FSO/THz Backhaul Link
Authors:
Pranay Bhardwaj,
Vedant Bansal,
Nikhil Biyani,
Shiv Shukla,
S. M. Zafaruddin
Abstract:
Establishing end-to-end connectivity of Internet of Things (IoT) network with the core for collecting sensing data from remote and hard-to-reach terrains is a challenging task. In this article, we analyze the performance of an IoT network integrated with wireless backhaul link for data collection. We propose a solution that involves a self-configuring protocol for aggregate node (AN) selection in…
▽ More
Establishing end-to-end connectivity of Internet of Things (IoT) network with the core for collecting sensing data from remote and hard-to-reach terrains is a challenging task. In this article, we analyze the performance of an IoT network integrated with wireless backhaul link for data collection. We propose a solution that involves a self-configuring protocol for aggregate node (AN) selection in an IoT network, which sends the data packet to an unmanned aerial vehicle (UAV) over radio frequency (RF) channels. We adopt a novel hybrid transmission technique for wireless backhaul employing opportunistic selections combining (OSC) and maximal ratio combining (MRC) that simultaneously transmits the data packet on mmWave (mW), free space optical (FSO), and terahertz (THz) technologies to take advantage of their complementary characteristics. We employ the decode-and-forward (DF) protocol to integrate the IoT and backhaul links and provide physical layer performance assessment using outage probability and average bit-error-rate (BER) under diverse channel conditions. We also develop simplified expressions to gain a better understanding of the system's performance at high signal-to-noise ratio (SNR). We provide computer simulations to compare different wireless backhaul technologies under various channel and SNR scenarios and demonstrate the performance of the data collection using the integrated link.
△ Less
Submitted 10 June, 2023; v1 submitted 3 April, 2023;
originally announced April 2023.
-
Communication-Control Co-design in Wireless Edge Industrial Systems
Authors:
Mark Eisen,
Santosh Shukla,
Dave Cavalcanti,
Amit S. Baxi
Abstract:
We consider the problem of controlling a series of industrial systems, such as industrial robotics, in a factory environment over a shared wireless channel leveraging edge computing capabilities. The wireless control system model supports the offloading of computational intensive functions, such as perception workloads, to an edge server. However, wireless communications is prone to packet loss an…
▽ More
We consider the problem of controlling a series of industrial systems, such as industrial robotics, in a factory environment over a shared wireless channel leveraging edge computing capabilities. The wireless control system model supports the offloading of computational intensive functions, such as perception workloads, to an edge server. However, wireless communications is prone to packet loss and latency and can lead to instability or task failure if the link is not kept sufficiently reliable. Because maintaining high reliability and low latency at all times prohibits scalability due to resource limitations, we propose a communication-control co-design paradigm that varies the network quality of service (QoS) and resulting control actions to the dynamic needs of each plant. We further propose a modular learning framework to solve the complex learning task without knowledge of plant or communication models in a series of learning steps and demonstrate its effectiveness in learning resource-efficient co-design policies in a robotic conveyor belt task.
△ Less
Submitted 8 February, 2022;
originally announced February 2022.
-
Fast and Accurate Single-Image Depth Estimation on Mobile Devices, Mobile AI 2021 Challenge: Report
Authors:
Andrey Ignatov,
Grigory Malivenko,
David Plowman,
Samarth Shukla,
Radu Timofte,
Ziyu Zhang,
Yicheng Wang,
Zilong Huang,
Guozhong Luo,
Gang Yu,
Bin Fu,
Yiran Wang,
Xingyi Li,
Min Shi,
Ke Xian,
Zhiguo Cao,
Jin-Hua Du,
Pei-Lin Wu,
Chao Ge,
Jiaoyang Yao,
Fangwen Tu,
Bo Li,
Jung Eun Yoo,
Kwanggyoon Seo,
Jialei Xu
, et al. (13 additional authors not shown)
Abstract:
Depth estimation is an important computer vision problem with many practical applications to mobile devices. While many solutions have been proposed for this task, they are usually very computationally expensive and thus are not applicable for on-device inference. To address this problem, we introduce the first Mobile AI challenge, where the target is to develop an end-to-end deep learning-based d…
▽ More
Depth estimation is an important computer vision problem with many practical applications to mobile devices. While many solutions have been proposed for this task, they are usually very computationally expensive and thus are not applicable for on-device inference. To address this problem, we introduce the first Mobile AI challenge, where the target is to develop an end-to-end deep learning-based depth estimation solutions that can demonstrate a nearly real-time performance on smartphones and IoT platforms. For this, the participants were provided with a new large-scale dataset containing RGB-depth image pairs obtained with a dedicated stereo ZED camera producing high-resolution depth maps for objects located at up to 50 meters. The runtime of all models was evaluated on the popular Raspberry Pi 4 platform with a mobile ARM-based Broadcom chipset. The proposed solutions can generate VGA resolution depth maps at up to 10 FPS on the Raspberry Pi 4 while achieving high fidelity results, and are compatible with any Android or Linux-based mobile devices. A detailed description of all models developed in the challenge is provided in this paper.
△ Less
Submitted 17 May, 2021;
originally announced May 2021.
-
Design Of Fuzzy Logic Traffic Controller For Isolated Intersections With Emergency Vehicle Priority System Using MATLAB Simulation
Authors:
Mohit Jha,
Shailja Shukla
Abstract:
Traffic is the chief puzzle problem which every country faces because of the enhancement in number of vehicles throughout the world, especially in large urban towns. Hence the need arises for simulating and optimizing traffic control algorithms to better accommodate this increasing demand. Fuzzy optimization deals with finding the values of input parameters of a complex simulated system which resu…
▽ More
Traffic is the chief puzzle problem which every country faces because of the enhancement in number of vehicles throughout the world, especially in large urban towns. Hence the need arises for simulating and optimizing traffic control algorithms to better accommodate this increasing demand. Fuzzy optimization deals with finding the values of input parameters of a complex simulated system which result in desired output. This paper presents a MATLAB simulation of fuzzy logic traffic controller for controlling flow of traffic in isolated intersections. This controller is based on the waiting time and queue length of vehicles at present green phase and vehicles queue lengths at the other phases. The controller controls the traffic light timings and phase difference to ascertain sebaceous flow of traffic with least waiting time and queue length. In this paper, the isolated intersection model used consists of two alleyways in each approach. Every outlook has different value of queue length and waiting time, systematically, at the intersection. The maximum value of waiting time and vehicle queue length has to be selected by using proximity sensors as inputs to controller for the ameliorate control traffic flow at the intersection. An intelligent traffic model and fuzzy logic traffic controller are developed to evaluate the performance of traffic controller under different pre-defined conditions for oleaginous flow of traffic. Additionally, this fuzzy logic traffic controller has emergency vehicle siren sensors which detect emergency vehicle movement like ambulance, fire brigade, Police Van etc. and gives maximum priority to him and pass preferred signal to it.
△ Less
Submitted 5 May, 2014;
originally announced May 2014.
