SSD (Single Shot Multibox Detector) is one of the best object detection algorithms with both high accuracy and fast speed. However, SSD's feature pyramid detection method makes it hard to fuse the features from different scales. In this paper, we proposed FSSD (Feature Fusion Single Shot Multibox Detector), an enhanced SSD with a novel and lightweight feature fusion module which can improve the performance significantly over SSD with just a little speed drop. In the feature fusion module, features from different layers with different scales are concatenated together, followed by some down-sampling blocks to generate new feature pyramid, which will be fed to multibox detectors to predict the final detection results. On the Pascal VOC 2007 test, our network can achieve 82.7 mAP (mean average precision) at the speed of 65.8 FPS (frame per second) with the input size 300×300 using a single Nvidia 1080Ti GPU. In addition, our result on COCO is also better than the conventional SSD with a large margin. Our FSSD outperforms a lot of state-of-the-art object detection algorithms in both aspects of accuracy and speed. Code is available at https://github.com/ lzx1413/CAFFE_SSD/tree/fssd.

No abstract

Binocular vision calibration is of great importance in 3D machine vision measurement. With respect to binocular vision calibration, the nonlinear optimization technique is a crucial step to improve the accuracy. The existing optimization methods mostly aim at minimizing the sum of reprojection errors for two cameras based on respective 2D image pixels coordinate. However, the subsequent measurement process is conducted in 3D coordinate system which is not consistent with the optimization coordinate system. Moreover, the error criterion with respect to optimization and measurement is different. The equal pixel distance error in 2D image plane leads to diverse 3D metric distance error at different position before the camera. To address these issues, we propose a precise calibration method for binocular vision system which is devoted to minimizing the metric distance error between the reconstructed point through optimal triangulation and the ground truth in 3D measurement coordinate system. In addition, the inherent epipolar constraint and constant distance constraint are combined to enhance the optimization process. To evaluate the performance of the proposed method, both simulative and real experiments have been carried out and the results show that the proposed method is reliable and efficient to improve measurement accuracy compared with conventional method.

The potential of 23 superhalogen anions of halogen-free structures as high-performance electrolytes of Li-ion batteries is theoretically explored here. According to high-level ab initio results at the CCSD(T) level, eight candidates, obeying the Wade-Mingos rule, should be capable of forming electrolytes, which are better than the currently used commercial products. When comparing different methods, MP2 was found to be in good agreement with CCSD(T) in the calculation of ΔE and ΔE, which are parameters describing the performance of potential electrolytes. Thus, MP2 represents a good choice for such calculations, particularly for large potential electrolyte systems wherein CCSD(T) calculations are actually impractical. The five functionals selected here (ωB97XD, B2GP-PLYP, B2K-PLYP, B2T-PLYP and B3LYP) are also capable of reproducing the variational trends of the relative values of different structures at the CCSD(T) level. However, the actual DFT values of ΔE are usually different from those of CCSD(T) by more than 1 eV. These significant deviations may be a problem when accurate ΔE values are required.

No abstract