Alibaba Group Holding Limited

Hangzhou Hikvision Digital Technology Co., Ltd.

Xing Shi

Wei Chu

Bin Guo

Feng Xie

Xudong Mao

Mengli Cheng

Wenmeng Zhou

Qiming Wan

Mingling Zhu

Ke Zhang

Shukun Xie

The present disclosure provides multi-sampling model training methods and devices. One exemplary training method includes: performing multi-sampling on samples to obtain a training set and a validation set in each sampling; using the training set and the validation set obtained in each sampling as a group, and performing model training and obtaining a trained model using the training set in each group; evaluating the trained model using the training set and the validation set in each group separately; eliminating or retaining the trained model based on the evaluation results and a predetermined elimination criterion; obtaining prediction results of the samples using retained models; and obtaining a final model by performing combined model training on the retained models using the prediction results. The final model obtained using embodiments of the present disclosure can be more robust and stable, and can provide more accurate prediction results, thus greatly improving efficiency of modeling.

Multi-sampling model training method and device

A character recognition method and apparatus are disclosed, relate to the field of image recognition technologies, reduce a calculation time, and increase character recognition efficiency. The method includes: obtaining picture data (); performing calculation on the picture data by using calculation layers shared by a first classifier and a second classifier, to obtain a first result, where the first classifier is a classifier for recognizing a specific character in the picture data, and the second classifier is a classifier for recognizing whether the picture data is a character picture (); performing calculation by bringing the first result into remaining calculation layers other than the shared calculation layers in the first classifier, to obtain a first probability corresponding to each character (); performing calculation by bringing the first result into remaining calculation layers other than the shared calculation layers in the second classifier, to obtain a second probability (); calculating, based on the first probability and the second probability, a confidence value that the picture data is recognized as each character (); and outputting a character recognition result based on the confidence ().

Character recognition method and device

The embodiments of the present application provide a method, apparatus and system for processing an image of a logistics parcel, which can capture an image with a barcode in a proper position. The method comprises: identifying a barcode in a Nth image to acquire position of the barcode in the Nth image, and placing the position into a queue; predicting a position that the barcode would locate in a (N+1)th image to be captured; identifying a barcode in the (N+1)th image to acquire position of the barcode relative to the (N+1)th image; determining whether the position of the barcode in the (N+1)th image matches with the predicted position; if so, placing the information of the barcode in the (N+1)th image into the queue; and after the parcel leaves the field of view, selecting, according to a preset optimal position of the barcode in an image and the information of the barcode placed into the queue, an optimal image from all images captured for the parcel. The present application is adapted to acquire logistic parcel information.

Logistics parcel picture processing method, device and system

Disclosed are a method and a system for video stitching. After it is determined that initial mounting of cameras in a M&#xd7;N video stitching scene is completed, a reference image is selected from M&#xd7;N images corresponding to M&#xd7;N cameras; where both M and N are positive integers, at least one of M and N is larger than 1, and a rotating mirror is set in front of a camera lens of each camera; an image whose position does not meet a requirement among respective images except for the selected reference image is adjusted by controlling a rotating mirror corresponding to the image to rotate according to the selected reference image; and video stitching is performed according to an adjustment result.

Method and system for video stitching

Growing community of inventors

Hangzhou, China

Xing Shi