Detection model training method and device, target detection method and device and storage medium

Abstract

The embodiment of the invention relates to a detection model training method, a target detection method and device and a storage medium. The method comprises the following steps: generating second laser data according to first laser data; wherein the line number of the first laser data is higher than that of the second laser data; training a knowledge distillation student network based on the second laser data, the knowledge distillation teacher network and a preset loss function to obtain a target student network; wherein the knowledge distillation teacher network is related to the first laser data, and the structures of the knowledge distillation teacher network and the knowledge distillation student network are the same. According to the method, under the application scene of adopting the low-line-number laser radar, the public large-scale high-line-number data set is fully utilized, the training data set is prevented from being collected again, the target student network can learn the parameters in the knowledge distillation teacher network by adopting the knowledge distillation mode, and the performance of the target student network is improved.

Description

technical field [0001] The embodiments of the present application relate to the field of computer technology, and in particular, to a detection model training method, a target detection method, an apparatus, and a storage medium. Background technique [0002] As we all know, lidar sensors are widely used in the mapping, positioning and environmental perception of autonomous bodies (such as robots). High-line-count lidar can provide accurate three-dimensional information, but its price is also expensive. Therefore, in some lower-cost products, low-line-count lidars are usually deployed, and there is a problem of domain differences caused by the line-count. However, most of the existing public datasets are collected by high-line-count lidar. Then, for the application scenario of low-line-count lidar, in order to avoid repeatedly re-collecting the dataset in this application scenario, how to use high-line-count lidar? It is a technical problem to be solved urgently by those s...

AI Technical Summary

Problems solved by technology

Therefore, in some lower-cost products, low-line-number lidar is usually deployed, and there is a problem of domain difference caused by the number of lines.

Images