Zhenyu Li

Andrey Ignatov, Grigory Malivenko, Radu Timofte et al. · tencent-ai

Various depth estimation models are now widely used on many mobile and IoT devices for image segmentation, bokeh effect rendering, object tracking and many other mobile tasks. Thus, it is very crucial to have efficient and accurate depth estimation models that can run fast on low-power mobile chipsets. In this Mobile AI challenge, the target was to develop deep learning-based single image depth estimation solutions that can show a real-time performance on IoT platforms and smartphones. For this, the participants used a large-scale RGB-to-depth dataset that was collected with the ZED stereo camera capable to generated depth maps for objects located at up to 50 meters. The runtime of all models was evaluated on the Raspberry Pi 4 platform, where the developed solutions were able to generate VGA resolution depth maps at up to 27 FPS while achieving high fidelity results. All models developed in the challenge are also compatible with any Android or Linux-based mobile devices, their detailed description is provided in this paper.

Baichen Li, Scott R. Downen, Quan Dong et al.

Non-invasive continuous alcohol monitoring has potential applications in both population research and in clinical management of acute alcohol intoxication or chronic alcoholism. Current wearable monitors based on transdermal alcohol content (TAC) sensing are relatively bulky and have limited quantification accuracy. Here we describe the development of a discreet wearable transdermal alcohol (TAC) sensor in the form of a wristband or armband. This novel sensor can detect vapor-phase alcohol in perspiration from 0.09 ppm (equivalent to 0.09 mg/dL sweat alcohol concentration at 25 °C under Henry's Law equilibrium) to over 500 ppm at one-minute time resolution. The TAC sensor is powered by a 110 mAh lithium battery that lasts for over 7 days. In addition, the sensor can function as a medical "internet-of-things" (IoT) device by connecting to an Android smartphone gateway via Bluetooth Low Energy (BLE) and upload data to a cloud informatics system. Such wearable IoT sensors may enable large-scale alcohol-related research and personalized management. We also present evidence suggesting a hypothesis that perspiration rate is the dominant factor leading to TAC measurement variabilities, which may inform more reproducible and accurate TAC sensor designs in the future.