Academia Sinica

About: Academia Sinica is a facility organization based out in Taipei, Taiwan. It is known for research contribution in the topics: Population & Gene. The organization has 52086 authors who have published 65998 publications receiving 1728114 citations. The organization is also known as: Central Research Academy.

High-speed high-security signatures

Abstract: This paper shows that a $390 mass-market quad-core 2.4GHz Intel Westmere (Xeon E5620) CPU can create 109000 signatures per second and verify 71000 signatures per second on an elliptic curve at a 2128 security level. Public keys are 32 bytes, and signatures are 64 bytes. These performance figures include strong defenses against software side-channel attacks: there is no data flow from secret keys to array indices, and there is no data flow from secret keys to branch conditions.

Scale-invariant quantum anomalous Hall effect in magnetic topological insulators beyond the two-dimensional limit.

Abstract: We investigate the quantum anomalous Hall effect (QAHE) and related chiral transport in the millimeter-size ${({\mathrm{Cr}}_{0.12}{\mathrm{Bi}}_{0.26}{\mathrm{Sb}}_{0.62})}_{2}{\mathrm{Te}}_{3}$ films. With high sample quality and robust magnetism at low temperatures, the quantized Hall conductance of ${e}^{2}/h$ is found to persist even when the film thickness is beyond the two-dimensional (2D) hybridization limit. Meanwhile, the Chern insulator-featured chiral edge conduction is manifested by the nonlocal transport measurements. In contrast to the 2D hybridized thin film, an additional weakly field-dependent longitudinal resistance is observed in the ten-quintuple-layer film, suggesting the influence of the film thickness on the dissipative edge channel in the QAHE regime. The extension of the QAHE into the three-dimensional thickness region addresses the universality of this quantum transport phenomenon and motivates the exploration of new QAHE phases with tunable Chern numbers. In addition, the observation of scale-invariant dissipationless chiral propagation on a macroscopic scale makes a major stride towards ideal low-power interconnect applications.

Scaled-YOLOv4: Scaling Cross Stage Partial Network

Abstract: We show that the YOLOv4 object detection neural network based on the CSP approach, scales both up and down and is applicable to small and large networks while maintaining optimal speed and accuracy. We propose a network scaling approach that modifies not only the depth, width, resolution, but also structure of the network. YOLOv4-large model achieves state-of-the-art results: 55.4% AP (73.3% AP50) for the MS COCO dataset at a speed of 15 FPS on Tesla V100, while with the test time augmentation, YOLOv4-large achieves 55.8% AP (73.2 AP50). To the best of our knowledge, this is currently the highest accuracy on the COCO dataset among any published work. The YOLOv4-tiny model achieves 22.0% AP (42.0% AP50) at a speed of 443 FPS on RTX 2080Ti, while by using TensorRT, batch size = 4 and FP16-precision the YOLOv4-tiny achieves 1774 FPS.