Search

Place recognition is an essential spatial intelligence capability for robots to understand and navigate the world. However, recognizing places in natural environments remains a challenging task for robots because of resource limitations and changing environments. In contrast, humans and animals can robustly and efficiently recognize hundreds of thousands of places in different conditions. Here, we report a brain-inspired general place recognition system, dubbed NeuroGPR, that enables robots to recognize places by mimicking the neural mechanism of multimodal sensing, encoding, and computing through a continuum of space and time. Our system consists of a multimodal hybrid neural network (MHNN) that encodes and integrates multimodal cues from both conventional and neuromorphic sensors. Specifically, to encode different sensory cues, we built various neural networks of spatial view cells, place cells, head direction cells, and time cells. To integrate these cues, we designed a multiscale liquid state machine that can process and fuse multimodal information effectively and asynchronously using diverse neuronal dynamics and bioinspired inhibitory circuits. We deployed the MHNN on Tianjic, a hybrid neuromorphic chip, and integrated it into a quadruped robot. Our results show that NeuroGPR achieves better performance compared with conventional and existing biologically inspired approaches, exhibiting robustness to diverse environmental uncertainty, including perceptual aliasing, motion blur, light, or weather changes. Running NeuroGPR as an overall multi–neural network workload on Tianjic showcases its advantages with 10.5 times lower latency and 43.6% lower power consumption than the commonly used mobile robot processor Jetson Xavier NX. [ABSTRACT FROM AUTHOR]

Mobile medical care is a hot issue in current medical research. Due to the inconvenience of going to hospital for fetal heart monitoring and the limited medical resources, real-time monitoring of fetal health on portable devices has become an urgent need for pregnant women, which helps to protect the health of the fetus in a more comprehensive manner and reduce the workload of doctors. For the feature acquisition of the fetal heart rate (FHR) signal, the traditional feature-based classification methods need to manually read the morphological features from the FHR curve, which is time-consuming and costly and has a certain degree of calibration bias. This paper proposes a classification method of the FHR signal based on neural networks, which can avoid manual feature acquisition and reduce the error caused by human factors. The algorithm will directly learn from the FHR data and truly realize the real-time diagnosis of FHR data. The convolution neural network classification method named "MKNet" and recurrent neural network named "MKRNN" are designed. The main contents of this paper include the preprocessing of the FHR signal, the training of the classification model, and the experiment evaluation. Finally, MKNet is proved to be the best algorithm for real-time FHR signal classification. [ABSTRACT FROM AUTHOR]

The effect of traffic flow prediction plays an important role in routing selection. Traditional traffic flow forecasting methods mainly include linear, nonlinear, neural network, and Time Series Analysis method. However, all of them have some shortcomings. This paper analyzes the existing algorithms on traffic flow prediction and characteristics of city traffic flow and proposes a road traffic flow prediction method based on transfer probability. This method first analyzes the transfer probability of upstream of the target road and then makes the prediction of the traffic flow at the next time by using the traffic flow equation. Newton Interior-Point Method is used to obtain the optimal value of parameters. Finally, it uses the proposed model to predict the traffic flow at the next time. By comparing the existing prediction methods, the proposed model has proven to have good performance. It can fast get the optimal value of parameters faster and has higher prediction accuracy, which can be used to make real-time traffic flow prediction. [ABSTRACT FROM AUTHOR]

With the development of technology, the network structure has changed a lot. Many people regard the Internet of Things as the next-generation network structure, which means all the embedded devices can communicate with each other directly. However, some problems remain in IoT before it can be applied in a large scale. Blockchain, which has become a hot research topic in recent years, may be one of the solutions. However, currently, the transaction speed of blockchain is still a disadvantage compared to traditional transaction methods. This paper focuses on to implement a high-performance blockchain platform. After investigation of the current blockchain consensus algorithm and blockchain architecture, we propose: (1) an improved blockchain consensus algorithm, which is implemented based on the mortgage model instead of probability model; (2) a cross-chain protocol with transverse expansion capacity, which would support the message transmission among chains; (3) a high-performance cross-chain blockchain network structure, which could handle more than 1000 transactions per second per chain by verification. Experiments have been carried out, and shown that the cross-chain blockchain network structure we provided is feasible to meet the requirement of large-scale distributed IoT applications. [ABSTRACT FROM AUTHOR]

Trace metal contamination in water and bioaccumulation in aquatic organisms are human health risks of increasing concern. However, the bioaccumulation of trace metals in the organs of the mussel Cristaria plicata in Dongting Lake, China and the human health risks of mussel consumption are largely unknown. We investigated the concentrations of 15 trace metals and metalloids in surface water, sediments, and C. plicata organs (foot, gill, mantle, and visceral mass) and quantified the bioaccumulation and human health risk of these trace metals in specimens collected from Dongting Lake. The concentrations of most metals in surface water exceeded previously published background values. In contrast, the concentrations of most metals in sediments showed a decreasing trend. Overall, the metal concentrations in the gill and visceral masses of C. plicata were higher than those in the foot and mantle, and higher bioaccumulation capacities were observed for essential metals than for nonessential metals. The mean concentrations of the trace elements Zn, Pb, Cd, As, Cu, and Cr in C. plicata foot samples were lower than the threshold values established by international and Chinese organizations. The estimated daily intake (EDI) values of the essential metal Mn in C. plicata foot was higher than the recommended tolerable daily intake (TDI) values for juveniles. Only Mn for juveniles and As for both juveniles and adults may pose noncarcinogenic health risks through foot consumption. The hazard index (HI) values for adults and juveniles were higher than 1, suggesting significant risks of noncarcinogenic effects to humans by exposure to multiple metals. [Display omitted] [ABSTRACT FROM AUTHOR]

Recent advances in artificial intelligence have enhanced the abilities of mobile robots in dealing with complex and dynamic scenarios. However, to enable computationally intensive algorithms to be executed locally in multitask robots with low latency and high efficiency, innovations in computing hardware are required. Here, we report TianjicX, a neuromorphic computing hardware that can support true concurrent execution of multiple cross-computing-paradigm neural network (NN) models with various coordination manners for robotics. With spatiotemporal elasticity, TianjicX can support adaptive allocation of computing resources and scheduling of execution time for each task. Key to this approach is a high-level model, "Rivulet," which bridges the gap between robotic-level requirements and hardware implementations. It abstracts the execution of NN tasks through distribution of static data and streaming of dynamic data to form the basic activity context, adopts time and space slices to achieve elastic resource allocation for each activity, and performs configurable hybrid synchronous-asynchronous grouping. Thereby, Rivulet is capable of supporting independent and interactive execution. Building on Rivulet with hardware design for realizing spatiotemporal elasticity, a 28-nanometer TianjicX neuromorphic chip with event-driven, high parallelism, low latency, and low power was developed. Using a single TianjicX chip and a specially developed compiler stack, we built a multi-intelligent-tasking mobile robot, Tianjicat, to perform a cat-and-mouse game. Multiple tasks, including sound recognition and tracking, object recognition, obstacle avoidance, and decision-making, can be concurrently executed. Compared with NVIDIA Jetson TX2, latency is substantially reduced by 79.09 times, and dynamic power is reduced by 50.66%.