Neural Inference Search for Multiloss Segmentation Models.

Semantic segmentation is vital for many emerging surveillance applications, but current models cannot be relied upon to meet the required tolerance, particularly in complex tasks that involve multiple classes and varied environments. To improve performance, we propose a novel algorithm, neural inference search (NIS), for hyperparameter optimization pertaining to established deep learning segmentation models in conjunction with a new multiloss function. It incorporates three novel search behaviors, i.e., Maximized Standard Deviation Velocity Prediction, Local Best Velocity Prediction, and n -dimensional Whirlpool Search. The first two behaviors are exploratory, leveraging long short-term memory (LSTM)-convolutional neural network (CNN)-based velocity predictions, while the third employs n -dimensional matrix rotation for local exploitation. A scheduling mechanism is also introduced in NIS to manage the contributions of these three novel search behaviors in stages. NIS optimizes learning and multiloss parameters simultaneously. Compared with state-of-the-art segmentation methods and those optimized with other well-known search algorithms, NIS-optimized models show significant improvements across multiple performance metrics on five segmentation datasets. NIS also reliably yields better solutions as compared with a variety of search methods for solving numerical benchmark functions.

Effects of collision warning characteristics on driving behaviors and safety in connected vehicle environments.

With the emerging connected vehicle (CV) technologies, a novel in-vehicle omni-direction collision warning system (OCWS) is developed. For example, vehicles approaching from different directions can be detected, and advanced collision warnings caused by vehicles approaching from different directions can be provided. Effectiveness of OCWS in reducing crash and injury related to forward, rear-end and lateral collision is recognized. However, it is rare that the effects of collision warning characteristics including collision types and warning types on micro-level driver behaviors and safety performance is assessed. In this study, variations in drivers' responses among different collision types and between visual only and visual plus auditory warnings are examined. In addition, moderating effects by driver characteristics including drivers' demographics, years of driving experience, and annual driving distance are also considered. An in-vehicle human-machine interface (HMI) that can provide both visual and auditory warnings for forward, rear-end, and lateral collisions is installed on an instrumented vehicle. 51 drivers participate in the field tests. Performance indicators including relative speed change, time taken to accelerate/decelerate, and maximum lateral displacement are adopted to reflect drivers' responses to collision warnings. Then, generalized estimation equation (GEE) approach is applied to examine the effects of drivers' characteristics, collision type, warning type and their interaction on the driving performance. Results indicate that age, year of driving experience, collision type, and warning type can affect the driving performance. Findings should be indicative to the optimal design of in-vehicle HMI and thresholds for the activation of collision warnings that can increase the drivers' awareness to collision warnings from different directions. Also, implementation of HMI can be customized with respect to individual driver characteristics.

3D-DFM: Anchor-Free Multimodal 3-D Object Detection With Dynamic Fusion Module for Autonomous Driving.

Recent advances in cross-modal 3D object detection rely heavily on anchor-based methods, and however, intractable anchor parameter tuning and computationally expensive postprocessing severely impede an embedded system application, such as autonomous driving. In this work, we develop an anchor-free architecture for efficient camera-light detection and ranging (LiDAR) 3D object detection. To highlight the effect of foreground information from different modalities, we propose a dynamic fusion module (DFM) to adaptively interact images with point features via learnable filters. In addition, the 3D distance intersection-over-union (3D-DIoU) loss is explicitly formulated as a supervision signal for 3D-oriented box regression and optimization. We integrate these components into an end-to-end multimodal 3D detector termed 3D-DFM. Comprehensive experimental results on the widely used KITTI dataset demonstrate the superiority and universality of 3D-DFM architecture, with competitive detection accuracy and real-time inference speed. To the best of our knowledge, this is the first work that incorporates an anchor-free pipeline with multimodal 3D object detection.

Knowledge Implementation and Transfer With an Adaptive Learning Network for Real-Time Power Management of the Plug-in Hybrid Vehicle.

Essential decision-making tasks such as power management in future vehicles will benefit from the development of artificial intelligence technology for safe and energy-efficient operations. To develop the technique of using neural network and deep learning in energy management of the plug-in hybrid vehicle and evaluate its advantage, this article proposes a new adaptive learning network that incorporates a deep deterministic policy gradient (DDPG) network with an adaptive neuro-fuzzy inference system (ANFIS) network. First, the ANFIS network is built using a new global K-fold fuzzy learning (GKFL) method for real-time implementation of the offline dynamic programming result. Then, the DDPG network is developed to regulate the input of the ANFIS network with the real-world reinforcement signal. The ANFIS and DDPG networks are integrated to maximize the control utility (CU), which is a function of the vehicle's energy efficiency and the battery state-of-charge. Experimental studies are conducted to testify the performance and robustness of the DDPG-ANFIS network. It has shown that the studied vehicle with the DDPG-ANFIS network achieves 8% higher CU than using the MATLAB ANFIS toolbox on the studied vehicle. In five simulated real-world driving conditions, the DDPG-ANFIS network increased the maximum mean CU value by 138% over the ANFIS-only network and 5% over the DDPG-only network.

Investigation on the Cooperative Grasping Capabilities of Human Thumb and Index Finger.

The maximum cooperative grasping mass and diameter of the human thumb and index finger were investigated by 7560 grasp-release trials on various masses of solid cylinders and various sizes of rings. The maximum grasping mass of the participants' thumb-index finger depended on gender, age and the sum of thumb-index finger lengths (P < 0.05), but not on the hand-used and ratio of index finger to thumb length (P > 0.05). The maximum grasping diameter of the participants' thumb-index finger depended on the age, sum of thumb-index finger lengths and ratio of index finger to thumb length (P < 0.05), but not on the gender and hand-used (P > 0.05). There was a non-linear regression model for the dependence of the maximum grasping mass on gender, age and the sum of thumb-index finger lengths and another non-linear regression model for the dependence of the maximum grasping diameter on the age, sum of thumb-index finger lengths and ratio of index finger to thumb length. Two regression models were useful in the optimal size design of robotic hands intending to replicate thumb-index finger grasping ability. This research can help to define not only a reasonable grasp mass and size for a bionic robotic hand, but also the requirements for hand rehabilitation.