1,720,971 research outputs found
Benchmarking Microcontrollers with Ultra-Low Resolution Images Classification
No full textThe advancement of edge devices equipped with specialized hardware accelerators, data caches, and microcontroller units (MCUs) has brought the deployment and execution of Deep Neural Network (DNN) models closer to users and real-world sensor systems. This paper explores the potential of various specialized MCUs across four real-world applications (waste classification, presence detection, miniature robot detection, and sign language interpretation). We evaluate three well-known MCU (STM32H7, Arduino, and MAX78000) comparing their inference time and power/energy consumption on four ultra-low resolution image-classification datasets with varying input and task complexities. Our findings indicate that all MCUs deliver excellent performance in terms of inference time and energy consumption for real-time applications, with the MAX78000 outperforming the others in all the metrics considered, thanks to the use of a DNN accelerator
Building a Pipeline for Efficient Production of Synthetic Datasets for Improving RL in Automated Driving
No full textOnline deep reinforcement learning training poses challenges due to its length and instability, despite the development of learning algorithms targeted to overcome these issues. Offline learning has emerged as a potential solution, but it reintroduces the issue of dataset production, which is resource-consuming and challenging even in simulation environments. This paper investigates efficient dataset creation for offline learning in the context of automated driving. Our proposed solution is a pipeline based on the CARLA simulator, which offers a wide variety in terms of car models, weather conditions, and environments. The pipeline aims to produce high-quality datasets for pre-training, training, and fine-tuning models, targeting improved training speed and reduced divergence. By leveraging CARLA's level of realism, we address the resource-intensive nature of dataset production, providing researchers and car manufacturers with a valuable tool for advancing the development of robust automated driving systems
Quantization of MobileNetV2 for Resource-Constrained Microcontrollers
No full textThe proliferation of IoT and edge computing devices demands for efficient deployment strategies due to their limited computational capabilities and energy constraints. This paper investigates the application of deep learning models in such resource-constrained environments, focusing on the quantization of the MobileNetV2 architecture. We evaluate three primary quantization techniques: dynamic quantization, static quantization, and quantization-aware training. Using a subset of the Vehicle Make and Model Recognition dataset, specifically the Most Stolen Vehicles in the US in 2017, we compare our results with a previous study, highlighting the advancements achieved through MobileNetV2 and its quantization process. Deployment on two STM32 boards, L4 and H7 series, demonstrates the effectiveness of the quantized MobileNetV2 model in achieving efficient, low-power, and low-latency execution on low-power MCUs
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Low-Cost, Edge-Cloud, End-to-End System Architecture for Human Activity Data Collection
No full textResearch in the Internet of Things (IoT) have paved the way to a new generation of applications and services that collect huge quantities of data from the field and do a significant part of the processing on the edge. This requires availability of efficient and effective methodologies and tools for a workflow spanning from the edge to the cloud. This paper presents a generic, complete workflow and relevant system architecture for field data collection and analysis with a focus on the human physical activities. The data source is given by a low-cost embedded system that can be placed on the user body to collect heterogeneous data on the performed movements. The system features a 9 DoF IMU sensor, to ensure a high level of configurability, connected to a custom board equipped with a rechargeable battery for wireless data collection. Data are transmitted via Bluetooth Low Energy (BLE) to a smartphone/tablet app, which manages the data transfer to Measurify, a cloud-based open-source framework designed for building measurement-oriented applications. Results from a preliminary functional experiment confirm the ability of the proposed end-to-end system architecture to efficiently implement the whole targeted edge-cloud workflow
Performance Comparison of YOLOv8 and YOLOv10 for Traffic Light Detection on a Jetson Nano Board
No full textDetecting traffic lights is crucial for the development of advanced driver-assistance systems (ADAS). Accurate traffic light detection allows for timely responses to changing traffic conditions, thereby enhancing road safety and reducing accidents. However, achieving high detection accuracy with small model size and fast inference times poses significant challenges, particularly for deployment on resource-constrained devices. In this study, we perform a detailed performance comparison of YOLOv8 and YOLOv10 nano models for traffic light detection (TLD), specifically targeting deployment on an NVIDIA Jetson Nano board. The evaluation focuses on key metrics including mean average precision (mAP), inference speed, and computational efficiency under real-time constraints. YOLOv8 demonstrated slightly superior mAP, indicating better detection accuracy. In contrast, YOLOv10 exhibited faster inference speeds due to its architectural optimizations. This comparison underscores the trade-offs between model complexity and deployment feasibility in embedded systems, providing insights for selecting the appropriate model for TLD applications in resource-constrained environments. Further research is needed to explore additional datasets, particularly those containing traffic lights at night, and to apply quantized models on smaller edge devices
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
TinyML Acceleration with MAX78000
No full textThe advancement of edge devices equipped with specialized hardware accelerators has brought the deployment and execution of Deep Neural Network (DNN) models nearer to users and real-world sensor systems. This paper investigates the potential of the MAX78000 microcontroller in accelerating Tiny Machine Learning applications, which require real-time processing and low power consumption. We compare its performance against other platforms like the STM32H7 and Raspberry Pi 4, focusing on a case study involving the detection of miniature mobile robots using an ultra-low-resolution Time-of-Flight sensor. Despite slightly lower accuracy, the MAX78000 outperforms other platforms in terms of inference time, power, and energy consumption, making it a reliable choice for power-constrained applications
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
- …
