Voice activity detection (VAD) is essential for speech-driven applications, but remains far from perfect in noisy and resource-limited environments. Existing methods often lack robustness to noise, and their frame-wise classification losses are only loosely coupled with the evaluation metric of VAD. To address these challenges, we propose SincQDR-VAD, a compact and robust framework that combines a Sinc-extractor front-end with a novel quadratic disparity ranking loss. The Sinc-extractor uses learnable bandpass filters to capture noise-resistant spectral features, while the ranking loss optimizes the pairwise score order between speech and non-speech frames to improve the area under the receiver operating characteristic curve (AUROC). A series of experiments conducted on representative benchmark datasets show that our framework considerably improves both AUROC and $F_{2}$-Score, while using only $69 \%$ of the parameters compared to prior arts, confirming its efficiency and practical viability.

Although the Tree Structured Model (TSM) is proven effective for solving face detection, pose estimation and landmark localization in an unified model, its sluggish run time makes it unfavorable in practical applications, especially when dealing with cases of multiple faces. We propose the Regressive Tree Structure Model (RTSM) to improve the run-time speed and localization accuracy. The RTSM is composed of two component TSMs, the coarse TSM (c-TSM) and the refined TSM (r-TSM), and a Bilateral Support Vector Regressor (BSVR). The c-TSM is built on the low-resolution octaves of samples so that it provides coarse but fast face detection. The r-TSM is built on the mid-resolution octaves so that it can locate the landmarks on the face candidates given by the c-TSM and improve precision. The r-TSM based landmarks are used in the forward BSVR as references to locate the dense set of landmarks, which are then used in the backward BSVR to relocate the landmarks with large localization errors. The forward and backward regression goes on iteratively until convergence. The performance of the RTSM is validated on three benchmark databases, the Multi-PIE, LFPW and AFW, and compared with the latest TSM to demonstrate its efficacy.

Although the Tree Structured Model (TSM) is proven effective for face detection, pose estimation and landmark localization, its sluggish runtime makes it unfavorable in practical applications. We propose the Bilayer Tree Structure Model (BTSM) to improve the run-time speed while keeping the performance unchanged or slightly better. The BTSM is composed of two component TSMs, the coarse c-TSM and the refined r-TSM. The c-TSM is trained on low-resolution samples so that it can provide coarse but fast detection, The r-TSM is trained on mid-resolution samples so that it can locate precise part locations. The performance of the BTSM is validated on three benchmark databases, the Multi-PIE, LFPW and AFW, and compared with the latest TSM to demonstrate its efficacy.

Despite the globally recognized advantages of e-invoicing and various efforts to implement such systems, retailers and stores may still have difficulties in promoting purely paperless e-invoices due to the lack of a convenient and secure way for consumers to receive and retrieve the e-invoices. As such, paper-based invoices may still be issued along with e-invoices, contradicting an important benefit of e-invoicing — paper consumption reduction. Thanks to the advances in smart phones and Near Field Communication (NFC) technologies, e-invoices can be delivered via NFC-enabled smartpones, allowing consumers to examine the content immediately after transactions and to easily retrieve them later on. Still, an extra security mechanism is needed to ensure the integrity and non-repudiation of the content, as invoices may bear some value and thus become the target of a security attack. In this paper, we propose a secure NFC-based e-invoice remitting scheme using standard NFC P2P communications, and discuss how it fulfills major security requirements, including authenticity, integrity, and non-repudiation. The proposed system is also implemented and tested in Taiwan's e-invoicing system.

Employees should continue their learning activities to face the changing and more competitive business environment. Nowadays, information technology plays an important role of building an e-learning system. In particular, employees work in a collaborative business environment can take advantage of IT-based e- learning system to enhance their learning and share knowledge with others among the business alliance. In this environment, traditional on-job training programs require specific space and time for the tutors and learners that cause extra expenses and reduce learning motivation. Therefore, building a dynamic e-learning system through IT becomes an essential part of the employee training programs. It should provide timely knowledge for employees to complete their assigned tasks individually or collaboratively with partners. A structure with five modules of dynamically configuring appropriate learning objects for learners is proposed. An interactive mode of retrieving learning materials based on individual requirement is presented. Personalized information is then issued to employees and partners to improve problem-solving capability.