In light of Industry 4.0, rapid analysis and optimization of manufacturing processes are emerging as a vital demand of smart manufacturing factories. Ergonomics is an essential aspect of the ongoing screening of working conditions and a fundamental variable in Industry 4.0, as it calls for a flexible manufacturing system to strengthen the competitiveness of factories in the global market. This paper proposes a new augmented reality-based ergonomic evaluation platform: Augmented Reality-based Ergonomic Platform (ARE Platform). Introducing the Augmented Reality technology by superimposing virtual planning objects into the physically existing production environment. Utilizing the motion capture system collects data for a set of ergonomic indexes (RULA, OWAS, and NIOSH), accessibility and visibility verification. ARE platform could be used in the verification phase of smart manufacturing systems to evaluate the level of risk to workers’ bodies during operations in real-time. The platform reduces the time and economic cost of verification and satisfies the rapid response of ergonomic evaluation and feedback in the context of smart manufacturing. Finally, the developed ARE platform is validated in two rigorous automotive assembly cases in the laboratory. Meanwhile, the ergonomic assessment results are analysed and reported for a people-oriented manufacturing system.

Fingerprint presentation attack detection (FPAD) is essential in fingerprint identification systems. Noncontact methods such as fingerprint biometrics are becoming popular because they are not affected by skin conditions and there are no hygiene issues. However, most of the existing noncontact FPAD methods are supervised methods with poor generalizability and poor performance during events such as unseen presentation attacks (PAs). Moreover, easily overlooked frequency domain information contributes to the fingerprint antispoofing task. Therefore, we propose a wavelet-based memory-augmented autoencoder that fully utilizes the frequency domain information. Specifically, the model first decomposes the input image into high- and low-frequency information and extracts features separately. Subsequently, we propose a frequency complementary connection (FCC) module to realize the fusion and complementation of frequency domain information at the feature level. Moreover, a memory distance expansion loss is proposed to keep the memory module diverse. Experiments are conducted to verify the effectiveness of the method. The code of our model is available onhttps://github.com/SuperIOyht/WaveMemAE.