Baidu USA LLC

Baidu.com Times Technology (beijing) Co. Ltd.

Zhenyu Zhong

Tao Wei

Zhisheng Hu

Yulong Zhang

Kang Li

Xinyang Zhang

Yunhan Jia

Zhebin Qiu

Yueqiang Cheng

Wei Ping

Xing Li

Shengjian Guo

Provided are a data processing method and apparatus, an electronic device, and a medium, and relates to the technical field of computers, and in particular to the field of autonomous driving and intelligent transportation. An implementation is: obtaining risk information in driving scenario data of a driving scenario; determining a risk factor set based on the risk information, where the risk factor set includes a risk factor existing in the driving scenario data; and generating a simulated risk scenario based on the risk factor set, where the simulated risk scenario reflects at least one risk factor in the risk factor set.

Data processing

The disclosure describes various embodiments for detecting an unexpected control state of an autonomous driving system. According to an embodiment, an exemplary method of detecting an unexpected control state of an autonomous driving system include the operations of generating environmental data of a vehicle; determining, by the autonomous driving system, a first control state based on the environmental data of the vehicle; determining, by a reference model, a second control state based on the environmental data, wherein the reference model defines at least one scenario each corresponding to a plurality of expected control states and a state switching condition, and in each of the expected control states corresponding to the scenario, an action of the vehicle in the scenario obeys a traffic rule; and determining the unexpected control state of the autonomous driving system by comparing the first control state with the second control state.

Method and apparatus for detecting unexpected control state in autonomous driving system

The present disclosure provides an image processing method and apparatus, an electronic device and a storage medium, which relate to the field of computer technology, and more particularly to artificial intelligence technology including computer vision, deep learning and the like. The image processing method includes: recognizing the image to be processed to determine attribute information of each object included in the image to be processed; determining a target thermal image to be recognized according to the attribute information of each object and the image to be processed; reconstructing the target thermal image to generate a first reconstructed image; and determining whether the image to be processed includes an object of a preset class according to a difference between the first reconstructed image and the target thermal image.

Image processing method and apparatus, electronic device and storage medium

Embodiments of the disclosure disclose a system to mitigate against adversarial input samples for machine learning (ML)/artificial intelligence (AI) models. According to one embodiment, a system receives a query from a client for a ML service. The system calculates a similarity score for the query based on a number of prior queries received from the client, the similarity score representing a similarity between the received query and the prior queries. The system determines that the query is an adversarial query in response to determining that the similarity score is above a predetermined threshold.

System to mitigate against adversarial samples for ML and AI models

According to various embodiments, an end-to-end TTS framework can integrate a watermarking process into the training of the TTS framework, which enables watermarks to be imperceptible within a synthesized/cloned audio segment generated by the TTS framework. The watermarks added in such a matter are statistically undetectable to prevent authorized removal. According to an exemplary method of training the TTS framework, a TTS neural network model and a watermarking neural network mode in the TTS framework are trained in an end to end manner, with the watermarking being part of the optimization process of the TTS framework. During the training, neuron values of the TTS neural network model are adjusted based on training data to prepare one or more spaces for adding a watermark in a synthesized audio segment to be generated by the TTS framework.

Inaudible watermark enabled text-to-speech framework

In one embodiment, a system receives a first image captured by a capturing device of an ADV. The system applies an image transformation to the first image to generate a second image. The system applies an object detection model to the first and the second images to generate a first and a second output, respectively. The system calculates a similarity metric between the first and the second output. The system detects the first image as an adversarial sample if a temporal variation of the similarity metric between the first image and a prior image is above a threshold.

Detecting adversarial samples by a vision based perception system

Embodiments of the present disclosure disclose an authentication method, an electronic device, and a computer-readable program medium. A specific embodiment of the method comprises: acquiring multimedia input information of a user in response to receiving an authentication request from the user; parsing the multimedia input information to obtain a parsing result; authenticating the user based on the parsing result; and transmitting an authentication result to the user; wherein the multimedia input information includes human face video input information and voice input information of the user; and the parsing result includes at least two of: biometric feature information of the user, voice content information, and behavior feature information of the user. The embodiments may lower a risk that a hacker forges the information for authentication inputted by the user.

Authentication method, electronic device, and computer-readable program medium

Embodiments of the present disclosure disclose an identity authentication method, a terminal device, and a computer-readable storage medium. An embodiment of the method comprises: acquiring video data displaying a lip part mouth shape changing process of a user; parsing the video data to generate annotation information of a human face region displayed by each image in consecutive images constituting the video data and human face feature information corresponding to each image in the consecutive images; generating a lip part feature information sequence and a human face local part feature information sequence based on the generated annotation information and the consecutive images; and transmitting the generated human face feature information, the lip part feature information sequence, and the human face local part feature information sequence to a server, and receiving an identity verification result returned from the server. The embodiment enhances identity authentication security.

Growing community of inventors

Sunnyvale, CA, United States of America

Zhenyu Zhong