Beijing Baidu Netcom Science Technology Co., Ltd

Beijing Baidu Netcom Science and Technology Co., Ltd.

Baohua Lai

Yanjun Ma

Yi Liu

Dianhai Yu

Hao Song

Yuning Du

Ruoyu Guo

Chenxia Li

Qiwen Liu

Long Yan

Zeyu Chen

Yukun Zhang

Xueying Lv

Zhiliang Yu

Minghua Li

Hai Zeng

Yuying Hao

Zewu Wu

The present disclosure provides an image processing method and apparatus, and relates to the field of image processing, and in particular to the field of image annotation. An implementation is: obtaining an image to be processed including a target region to be annotated; in response to a first click on the target region, performing a first operation to expand a predicted region for the target region based on a click position of the first click; in response to a second click in a position where the predicted region exceeds the target region, performing a second operation to reduce the predicted region based on a click position of the second click; and in response to determining that a difference between the predicted region and the target region meets a preset condition, obtaining an outline of the predicted region to annotate the target region.

Image processing

Provided is a lightweight model training method, an image processing method, a device and a medium. The lightweight model training method includes: acquiring first and second augmentation probabilities and a target weight adopted in an e-th iteration; performing data augmentation on a data set based on the first and second augmentation probabilities respectively, to obtain first and second data sets; obtaining a first output value of a student model and a second output value of a teacher model based on the first data set; obtaining a third output value and a fourth output value based on the second data set; determining a distillation loss function, a truth-value loss function and a target loss function; training the student model based on the target loss function; and determining a first augmentation probability or target weight to be adopted in an (e+1)-th iteration in a case of e is less than E.

Lightweight model training method, image processing method, electronic device, and storage medium

The present disclosure provides a vision processing and model training method, device, storage medium and program product. A specific implementation solution is as follows: establishing an image classification network with the same backbone network as the vision model, performing a self-monitoring training on the image classification network by using an unlabeled first data set; initializing a weight of a backbone network of the vision model according to a weight of a backbone network of the trained image classification network to obtain a pre-training model, the structure of the pre-training model being consistent with that of the vision model, and optimize the weight of the backbone network by using real data set in a current computer vision task scenario, so as to be more suitable for the current computer vision task; then, training the pre-training model by using a labeled second data set to obtain a trained vision model.

Vision processing and model training method, device, storage medium and program product

Enclosed are a method and apparatus for operating a database. The method includes: performing, in response to determining a user indicated by a user identifier having a permission to operate on data targeted by a database operation statement, a digital signature on the database operation statement using a user private key to obtain signature information, generating transaction information according to the database operation statement, the signature information and the user identifier, and storing the generated transaction information into a local transaction information buffer pool; and selecting, in response to competing for a transaction information block recording permission using a preset consensus algorithm, transaction information in the local transaction information buffer pool to generate a block, parsing and executing a database operation statement in each piece of selected transaction information, concatenating the generated block to a local blockchain, and broadcasting the generated block to other database servers.

Method and apparatus for operating database

A distributed database transaction processing method and device based on a GPS atomic clock, the method includes: receiving a plurality of transaction requests by the first node server; determining a concurrency conflict between transaction requests received by the first node server; obtaining start time of each transaction in a case that the concurrency conflict exists, wherein the start time is local time of the second node server when the transaction request is sent; and local time of the second node server is synchronized with time of the GPS atomic clock server in an area which the second node server locates; processing all transactions by the first node server according to the sequence of the start time of each transaction. Throughput of distributed database system can be increased and an arrangement complexity of servers is reduced.

Distributed database transaction processing method, device based on GPS atomic clock server

Growing community of inventors

Beijing, China

Baohua Lai