Active Millimeter Wave (AMMW) imaging technique has been widely applied in the security industries due to its under-controlled privacy concerns and no health hazards. In fact, how to automatically and precisely detect the concealed objects on the human body in the AMMW images is one key issue in industrial security scanner systems. In this paper, we first investigate the deep learning based object detection approaches for the AMMW images, and then we develop a concealed object detection system for the security system in the airport. For our particular application, the concealed objects include several small items. However, most of the current deep learning based methods focus mainly on detecting large objects resulting in the unsatisfactory performance on small objects. To address this issue, we first skip the down-sampling of the last pooling layer and apply the dilated convolution for the following convolutional layers. Then, we propose a context embedding object detection network. Besides, we build a new large-scale dataset to benchmark the concealed object detection tasks in the AMMW images, which includes more than 58k AMMW images. With the proposed algorithm, we finally achieve 90.05% in detection rate and 9.73% in false alarm rate in our dataset.

Depressive disorders are the most common form of mental disorders in community and health care settings. Unfortunately, the treatment of Major Depressive Disorder (MDD) is far from satisfactory. Vagus nerve stimulation (VNS) is a relatively new and promising physical treatment for depressive disorders. One particularly appealing element of VNS is the long-term benefit in mood regulation. However, because this intervention involves surgery, perioperative risks, and potentially significant side effects, this treatment has been limited to those patients with treatment-resistant depression who have failed medication trials and exhausted established somatic treatments for major depression, due to intolerance or lack of response.

Insomnia inflicts mental burden and decreases physical productivity and affects life quality. Transcutaneous vagus nerve stimulation (ta‐VNS) may be an effective treatment option for insomnia. This study aims to evaluate the effect and safety of ta‐VNS and compare it with transcutaneous nonvagus nerve stimulation (tn‐VNS). A multicenter, randomized, clinical trial was conducted at 3 hospitals in China enrolling 72 insomnia participants from May 2016 to June 2017. Participants were randomly assigned (1 : 1) to receive 40 sessions of ta‐VNS or tn‐VNS treatment. 63 participants completed the trial. ta‐VNS treatment significantly decreased the Pittsburgh Sleep Quality Index score, Epworth Sleepiness Scale score, Flinders Fatigue Scale score, Hamilton Depression Scale score, and Hamilton Anxiety Scale score over 4 weeks compared with those of the baseline. Moreover, it also significantly increased the 36‐item Short‐Form Health Survey Questionnaire scores compared with that of the baseline. However, it did not show significant differences compared with tn‐VNS in changes of primary and secondary outcomes. The incidence of adverse events was low. ta‐VNS significantly relieved insomnia over 4 weeks. Moreover, it also alleviated fatigue and improved participants’ quality of life as well as other concomitant symptoms such as depression and anxiety. This trial is registered at Chinese Clinical Trial Registry (http://www.chictr.org.cn) with the registration number: ChiCTR-TRC-13003519.

Recently, clinical observations reported the potential benefit of vagus nerve stimulation (VNS) for pediatric epilepsy. Transcutaneous auricular vagus nerve stimulation (ta-VNS) is a newer non-invasive VNS, making it more accessible for treating pediatric epilepsy, yet there is limited clinical evidence for its effectiveness. A three-center, randomized, parallel, controlled trial will be carried out to evaluate whether ta-VNS improves pediatric epilepsy. Pediatric patients aged 2 to 14 years with epilepsy will be recruited and randomly assigned to transcutaneous auricular vagus nerve stimulation (ta-VNS) group, transcutaneous auricular non-vagus nerve stimulation (tan-VNS) group, and control group with a 1:1: sqrt(2) allocation, as per a computer generated randomization schedule stratified by study center using permuted blocks of random sizes. We will use Zelen’s design, in which randomization occurs before informed consent. Patients in the stimulation groups will receive tan-VNS or ta-VNS three times a day for 6 months. Patients in the control group will not be provided with any stimulation during the 6 months. The guardians of the patients are required to keep a detailed diary to record the data. Outcome assessment including seizure frequency, electroencephalogram (EEG), heart rate variability (HRV) analysis, quality of life (QOL) and adverse events will be made at baseline and 2, 4 and 6 months after ta-VNS initiation. The seizure frequency and adverse events will be followed up at 1 year and 1.5 years after ta-VNS initiation. Results of this trial will help clarify whether ta-VNS treatment is beneficial for pediatric patients, and will make clear whether the anticonvulsive effect of ta-VNS is correlated with the improvement of sympathovagal imbalance. Clinical Trials Identifier: NCT02004340 . Registration date: 13 November 2013.

During the past few years, significant progress has been made on cross-modal retrieval, benefiting from the development of deep neural networks. Meanwhile, the overall frameworks are becoming more and more complex, making the training as well as the analysis more difficult. In this paper, we provide a Rich Features Embedding (RFE) approach to tackle the cross-modal retrieval tasks in a simple yet effective way. RFE proposes to construct rich representations for both images and texts, which is further leveraged to learn the rich features embedding in the common space according to a simple hard triplet loss. Without any bells and whistles in constructing complex components, the proposed RFE is concise and easy to implement. More importantly, our RFE obtains state-of-the-art results on several popular benchmarks such as MS COCO and Flickr 30K. In particular, the image-to-text and text-to-image retrieval achieve 76.1% and 61.1% (R@1) on MS COCO, which outperform the others more than 3.4% and 2.3% respectively. We hope our RFE will serve as a solid baseline and help ease future research in cross-modal retrieval.