Chinese Academy of Medical Sciences

Beijing Kuangshi Technology Co., Ltd.

Megvii (beijing) Technology Co., Ltd.

Jiahui Yu

Yan Li

Qi Yin

Beifen Shen

Chunmei Hou

He Xiao

Gencheng Han

Zhi Zhao

Guojiang Chen

Jiannan Feng

Xingwei Jiang

Renxi Wang

The embodiments of the present invention provide training and construction methods and apparatus of a neural network for object detection, an object detection method and apparatus based on a neural network and a neural network. The training method of the neural network for object detection, comprises: inputting a training image including a training object to the neural network to obtain a predicted bounding box of the training object; acquiring a first loss function according to a ratio of the intersection area to the union area of the predicted bounding box and a true bounding box, the true bounding box being a bounding box of the training object marked in advance in the training image; and adjusting parameters of the neural network by utilizing at least the first loss function to train the neural network.

Method and apparatus for neural network training and construction and method and apparatus for object detection

The present invention provides a human Tim-3-Ig fusion protein which can block Tim-3 signal pathway, and said Ig fusion protein comprises Tim-3 protein, human Ig fragment, and the linking sequence therebetween. In the present invention, a human Tim-3-Ig gene is obtained by an artificial synthesis process; an expression vector containing the Tim-3-Ig gene is constructed; and the prepared Ig fusion protein is tested in an expression verification experiment, a binding activity experiment, a blocking activity experiment in different cell lines, and in vivo experiments in mice. The human Tim-3-Ig fusion protein prepared in the present invention can be used to treat immunological diseases caused by high expression of Tim-3.

Human Tim-3 fusion protein capable of blocking Tim-3 signaling pathway

Growing community of inventors

Beijing, China

Jiahui Yu