The computation and storage requirements for Deep Neural Networks (DNNs) are usually high. This issue limits their deployability on ubiquitous computing devices such as smart phones, wearables and autonomous drones. In this paper, we propose ternary neural networks (TNNs) in order to make deep learning more resource-efficient. We train these TNNs using a teacher-student approach based on a novel, layer-wise greedy methodology. Thanks to our two-stage training procedure, the teacher network is still able to use state-of-the-art methods such as dropout and batch normalization to increase accuracy and reduce training time. Using only ternary weights and activations, the student ternary network learns to mimic the behavior of its teacher network without using any multiplication. Unlike its {-1,1} binary counterparts, a ternary neural network inherently prunes the smaller weights by setting them to zero during training. This makes them sparser and thus more energy-efficient. We design a purpose-built hardware architecture for TNNs and implement it on FPGA and ASIC. We evaluate TNNs on several benchmark datasets and demonstrate up to 3.1× better energy efficiency with respect to the state of the art while also improving accuracy.
The real world human activity datasets are of great importance in development of novel machine learning methods for automatic recognition of human activities in smart environments. In this study, we present the details of ARAS (Activity Recognition with Ambient Sensing) human activity recognition datasets that are collected from two real houses with multiple residents during two months. The datasets contain the ground truth labels for 27 different activities. Each house was equipped with 20 binary sensors of different types that communicate wirelessly using the ZigBee protocol. A full month of information which contains the sensor data and the activity labels for both residents was gathered from each house, resulting in a total of two months data. In the paper, particularly, we explain the details of sensor selection, targeted activities, deployment of the sensors and the characteristics of the collected data and provide the results of our preliminary experiments on the datasets.
Human activity recognition and behavior monitoring in a home setting using wireless sensor networks (WSNs) provide a great potential for ambient assisted living (AAL) applications, ranging from health and wellbeing monitoring to resource consumption monitoring. However, due to the limitations of the sensor devices, challenges in wireless communication and the challenges in processing large amounts of sensor data in order to recognize complex human activities, WSN-based AAL systems are not effectively integrated in the home environment. Additionally, given the variety of sensor types and activities, selecting the most suitable set of sensors in the deployment is an important task. In order to investigate and propose solutions to such challenges, we introduce a WSN-based multimodal AAL system compatible for homes with multiple residents. Particularly, we focus on the details of the system architecture, including the challenges of sensor selection, deployment, networking and data collection and provide guidelines for the design and deployment of an effective AAL system. We also present the details of the field study we conducted, using the systems deployed in two different real home environments with multiple residents. With these systems, we are able to collect ambient sensor data from multiple homes. This data can be used to assess the wellbeing of the residents and identify deviations from everyday routines, which may be indicators of health problems. Finally, in order to elaborate on the possible applications of the proposed AAL system and to exemplify directions for processing the collected data, we provide the results of several human activity inference experiments, along with examples on how such results could be interpreted. We believe that the experiences shared in this work will contribute towards accelerating the acceptance of WSN-based AAL systems in the home setting.
Honey bees live in colonies of thousands of individuals, that not only need to collaborate with each other but also to interact intensively with their ecosystem. A small group of robots operating in a honey bee colony and interacting with the queen bee, a central colony element, has the potential to change the collective behavior of the entire colony and thus also improve its interaction with the surrounding ecosystem. Such a system can be used to study and understand many elements of bee behavior within hives that have not been adequately researched. We discuss here the applicability of this technology for ecosystem protection: A novel paradigm of a minimally invasive form of conservation through “Ecosystem Hacking”. We discuss the necessary requirements for such technology and show experimental data on the dynamics of the natural queen’s court, initial designs of biomimetic robotic surrogates of court bees, and a multi-agent model of the queen bee court system. Our model is intended to serve as an AI-enhanceable coordination software for future robotic court bee surrogates and as a hardware controller for generating nature-like behavior patterns for such a robotic ensemble. It is the first step towards a team of robots working in a bio-compatible way to study honey bees and to increase their pollination performance, thus achieving a stabilizing effect at the ecosystem level.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.