Method and apparatus for detecting object, method and apparatus for training neural network, and electronic device

An apparatus for detecting an object, comprising:

• a processor;
• and
• a memory for storing instructions executable by the processor;
• wherein the processor is configured to:
• predict to obtain a plurality of fusion feature maps from a to-be-processed image through a deep convolutional neural network for target area frame detection, wherein the deep convolutional neural network comprises a first sub-network and a second sub-network;
• the first sub-network has at least one down-sampling layer, and the second sub-network has at least one up-sampling layer;
• the fusion feature maps are obtained through a first feature map and a second feature map;
• and the first feature map is obtained from the first sub-network, and the second feature map is obtained from the second sub-network;
• and
• obtain target area frame data according to the plurality of fusion feature maps obtained by the fusion feature map prediction module.

A non-transitory computer-readable storage medium having stored therein instructions that, when executed by a processor, cause the processor to perform the operations of:

• predicting, by a deep convolutional neural network for target area frame detection, to obtain a plurality of fusion feature maps from a to-be-processed image, wherein the deep convolutional neural network comprises a first sub-network and a second sub-network;
• the first sub-network has at least one down-sampling layer, and the second sub-network has at least one up-sampling layer;
• the fusion feature maps are obtained through a first feature map and a second feature map;
• and the first feature map is obtained from the first sub-network, and the second feature map is obtained from the second sub-network;
• and
• obtaining target area frame data according to the plurality of fusion feature maps.