No abstract

Transcranial ultrasound stimulation (TUS; f < 1 MHz) is a promising approach to non-invasive brain stimulation. Transcranial magneto-acoustic stimulation (TMAS) is a technique of neuromodulation for regulating neuroelectric-activity utilizing a magnetic–acoustic coupling electric field generated by low-intensity ultrasound and magnetic fields. However, both techniques use the physical means of low-intensity ultrasound and can induce the response of the motor cortex. Therefore, it is necessary to distinguish the difference between the two techniques in the regulation of neural activity. This study is the first to quantify the amplitude and response latency of motor cortical electromyography (EMG) in mice induced by TMAS and TUS. The amplitude of EMG (2.73 ± 0.32 mV) induced by TMAS was significantly greater than that induced by TUS (2.22 ± 0.33 mV), and the EMG response latency induced by TMAS (101.25 ± 88.4 ms) was significantly lower than that induced by TUS (181.25 ± 158.4 ms). This shows that TMAS can shorten the response time of nerve activity and enhance the neuromodulation effect of TUS on the motor cortex. This provides a theoretical basis for revealing the physiological mechanisms of TMAS and the treatment of neuropsychiatric diseases using it.

The attention mechanism provides a sequential prediction framework for learning spatial models with enhanced implicit temporal consistency. In this work, we show a systematic design (from 2D to 3D) for how conventional networks and other forms of constraints can be incorporated into the attention framework for learning long-range dependencies for the task of pose estimation. The contribution of this paper is to provide a systematic approach for designing and training of attention-based models for the end-to-end pose estimation, with the flexibility and scalability of arbitrary video sequences as input. We achieve this by adapting temporal receptive field via a multi-scale structure of dilated convolutions. Besides, the proposed architecture can be easily adapted to a causal model enabling real-time performance. Any off-the-shelf 2D pose estimation systems, e.g. Our method achieves the state-of-the-art performance and outperforms existing methods by reducing the mean per joint position error to 33.4mm on Human 3.

No abstract

High-resolution aerial photographs of Arctic region 1 are a great source for different sea ice feature recognition, which 2 are crucial to validate, tune and improve climate models. Melt 3 ponds on the surface of melting Arctic sea ice are of particular 4 interest as they are sensitive and valuable indicators and are 5 proxy to the processes in the Arctic climate system. Manual 6 analysis of this remote sensing data is extremely difficult and 7 time-consuming due to the complex shapes and unpredictable 8 boundaries of the melt ponds, and that leads to the necessity 9 for automatizing the processes. In this study, we propose a 10 robust and efficient automatic method for melt pond region 11 segmentation and boundary extraction from high-resolution 12 aerial photographs. The proposed algorithm is based on a swin 13 transformer U-Net in which we introduce novel cross-channel 14 attention mechanisms into the decoder design. The framework 15 operates with optical data and allows for classifying imagery into 16 four classes: sea ice/snow, open water, melt pond, and submerged 17 ice. We use aerial photographs collected during the Healy-Oden 18 Trans Arctic Expedition (HO-TRAX) expedition over Arctic sea19 ice in the summer season of 2005 as test data. The experimental 20 results show that the proposed method is suitable for precise 21 automatic extraction of melt pond geometry and it can also be 22 extended for other optical data sources that involve melt ponds.