Fast and accurate single-stage target detection method and device

Abstract

The invention discloses a fast and accurate single-stage target detection method and a device thereof, which discard the improved method of complex bottom layer and high layer network structure, and enhance the coordination and unity between the multi-layer regression characteristic diagrams of the high layer network only by optimizing the main stream structure and adding a lightweight split structure (i.e. Shunt structure). The invention uses the high-level network to carry out the maximum pooling and asynchronous convolution operation on the characteristics of the output of the bottom network, which is beneficial to the transmission of the spatial related information while reducing the dimension of the characteristic map, can improve the problem of the missing of the spatial related information and promote the diversity and the difference of the characteristics. After the shunting structure is added and the mainstream structure is optimized, the average accuracy of target detection is 80.5%, is 3.3% higher than SSD and 1.9% higher than DSSD321. At the same time, the average processing speed of 30fps is achieved on a 1080ti graphics card.

Description

technical field [0001] The invention belongs to the technical field of target detection, and in particular relates to a fast and accurate single-stage target detection method and device. Background technique [0002] As a key technology in the field of computer vision, object detection has always been a challenging research hotspot. The Viola-Jones algorithm proposed by Viola and Jones in 2001 realized real-time face detection for the first time under the condition of limited computing resources. In 2005, the HOG pedestrian detector proposed by Dalal and Triggs expanded the detection field of Viola-Jones algorithm. The deformable part model DMP (Deformable Part based Model) proposed by Felzenszwalb et al. and its follow-up optimization algorithm have won the championship of the VOC Object Detection Challenge (The PASCAL Visual Object Classes Challenge) for three consecutive years, representing the detector based on hand-designed features at that time the highest level. Ho...

AI Technical Summary

Problems solved by technology

[0008] The technical problem to be solved by the present invention is: in order to solve the relatively independent problem of inter-layer regression calculation existing in the SSD series algorithm, while ensuring the real-time performance of target detection, and further improving the detection accuracy, a fast and accurate single-stage method is constructed based on SSD. The target detection method and device, abandoning the complex underlying and high-level network structure improvement method, only by optimizing the mainstream structure and adding a lightweight shunt structure (ie shunt structure) to enhance the coordination between multi-layer regression feature maps of the high-level network Sex and Unity

Images