Beijing Baidu Netcom Science Technology

The present disclosure provides a method for data processing, apparatus, electronic device, storage medium and computer program product, which relates to the field of artificial intelligence technology, specifically to the fields of intelligent cloud, network communication, large language models and other technologies. A method for data processing is applied to a bottom-layer switch in a multi-layer switch, wherein the multi-layer switch is configured to complete a target operation, the target operation includes a plurality of stage operations. The method includes: receiving a plurality of in-network computing requests sent by a top-layer switch in the multi-layer switch; wherein the plurality of in-network computing requests correspond to the plurality of stage operations one-to-one, and the plurality of in-network computing requests are sent by a current GPU to the top-layer switch; executing the plurality of stage operations in parallel for a plurality of GPUs in a current subgroup based on the plurality of in-network computing requests; wherein a target group to which the current GPU belongs is divided into a plurality of subgroups, and the current subgroup is a subgroup corresponding to the bottom-layer switch among the plurality of subgroups.

The present disclosure provides a data processing method, apparatus, device, medium and product, which relates to the field of artificial intelligence technology, specifically in the fields of intelligent cloud, network communication, and large language models. The data processing method is applied to a single-layer switch, where the single-layer switch is configured to complete a target operation, and the target operation includes multiple stage operations. The method includes: receiving multiple in-network computation requests sent by a current GPU, where the multiple in-network computation requests correspond to the multiple stage operations one by one; parallelly executing the multiple stage operations for multiple GPUs in a target group where the current GPU is located based on the multiple in-network computation requests.

The present disclosure provides an adaptive compensation method during video image encoding and a related apparatus, relating to the technical fields of cloud computing and video transcoding. The solution includes: acquiring an original pixel value array of an original pixel block to be encoded; acquiring a reconstructed pixel value array of a reconstructed pixel block obtained by reconstructing the original pixel block; determining an encoding category to which each pixel in the reconstructed pixel block belongs, and determining a respective target pixel value of each pixel simultaneously satisfying at least two optimization objectives; for each pixel under each encoding category, determining a first offset value according to an original pixel value of the pixel in the original pixel value array and a reconstructed pixel value of the pixel in the reconstructed pixel value array; determining a second offset value according to the target pixel value of the pixel and the reconstructed pixel value; determining an offset value interval of the pixel according to magnitudes of the first offset value and the second offset value; performing frequency statistics on integer offset values respectively included in the offset value intervals of all pixels under each encoding category, and determining an expected offset value of the corresponding encoding category based on a statistical result obtained; and compensating a corresponding reconstructed pixel value for each pixel according to the expected offset value corresponding to the encoding category to which the pixel belongs.

The present disclosure provides a joint lateral and longitudinal control method and apparatus for an autonomous vehicle, an electronic device, and a medium, and relates to the technical field of autonomous driving, in particular to the technical field of trajectory tracking and model predictive control. A solution for implementation includes: obtaining a lateral-longitudinal coupled state prediction model of an autonomous vehicle and a desired trajectory; obtaining a current state quantity and a current control quantity of the autonomous vehicle; solving for a target control quantity for trajectory tracking using a model predictive control algorithm based on the current state quantity, the current control quantity, the lateral-longitudinal coupled state prediction model, and the desired trajectory, where the model predictive control includes quadratic programming, and a cost function of the quadratic programming includes at least a lateral-longitudinal decoupled target error; and controlling the autonomous vehicle to perform the trajectory tracking based on the target control quantity.

Provided is a driving information prediction method, apparatus, and autonomous driving vehicle, which relate to the field of autonomous driving, especially to the field of artificial intelligence, and particularly to the technical fields of autonomous driving and intelligent transportation. The method includes: determining (S110) a first leader-follower relationship between a target vehicle and a first obstacle based on a motion parameter of the target vehicle at a current moment, path information of the target vehicle within a first time period, a motion parameter of the first obstacle at the current moment, and predicted path information of the first obstacle within the first time period; obtaining (S120) first predicted driving information based on the first leader-follower relationship; and determining (S130) first optimal driving information based on an evaluation result corresponding to the first predicted driving information.

The present disclosure provides a lane-level positioning method and apparatus, a device, a vehicle, and a medium, and relates to the field of computer technologies and specifically to the technical fields such as artificial intelligence, computer vision, intelligent vehicles, and autonomous vehicles. The method includes: obtaining map road information and visual perception road information of surroundings of a vehicle based on positioning information of the vehicle; determining a plurality of candidate lanes based on the positioning information; determining a topological recursion probability of the vehicle being in each of the candidate lanes by utilizing the map road information; determining a perception observation probability of the vehicle being in each of the candidate lanes by utilizing the visual perception road information; determining a positioning probability of the vehicle being in each of the candidate lanes by utilizing the positioning information; and determining a target lane the vehicle is in from the plurality of candidate lanes based on the topological recursion probability, the perception observation probability, and the positioning probability of the vehicle being in each of the candidate lanes.

The present disclosure provides an animal language translation method, an emotion recognition method, and a model training method and apparatus, which relates to the field of artificial intelligence technology, particularly to the field of machine learning technology, deep learning technology, natural language processing technology and the like. A specific implementation solution involves acquiring animal-related multimodal data; preprocessing the multimodal data to obtain fused multimodal data; recognizing current animal emotions based on the fused multimodal data to obtain animal emotion recognition results; performing semantic mapping and language translation on the emotion recognition results to translate animal languages into human languages to obtain language translation results.

Provided is a method and apparatus for automatically annotating an obstacle and related devices, relating to the field of artificial intelligence technology, and in particular, to technologies fields of autonomous driving, neural network, deep learning and the like. The method includes: optimizing (S101) a target parameter in a projection relationship based on a re-projection error, the projection relationship is used to project a target obstacle from a reference frame onto a frame to be optimized, and satisfies a constraint in which positions of the target obstacle in different frames are consistent in an obstacle coordinate system established according to the target obstacle; and determining (S102) a target pose of the target obstacle based on the optimized target parameter.

Provided is a method and an apparatus for controlling quality of service of a virtual machine, relating to the fields of cloud computing, virtualization, and big data. The method includes: storing (S101) an operation request in a target queue corresponding to a layer of a virtual machine cluster when an available token quantity in a token bucket of the layer is unable to meet a target token quantity required for the operation request; wherein tokens of the layer are periodically generated according to an upper limit of quality of service of the layer; deducting (S102) the target token quantity from an updated available token quantity when the available token quantity in the token bucket is updated to meet the target token quantity required for the operation request; and sending (S103) the operation request from the target queue to the layer for processing when deducting the target token quantity successfully.

The invention provides an image generation method and apparatus, an intelligent agent, an intelligent agent system and a storage medium, relates to the field of artificial intelligence technology, in particular to the field of computer vision, deep learning and large language models, and can be applied to an artificial intelligence generated content scene. The specific implementation solution includes: obtaining (101, 201, 301, 501, 601, 701) image generation requirement information; determining (102, 202, 502, 602, 702) a target image generation manner according to the image generation requirement information; querying (103, 305, 503) a first reference image based on the image generation requirement information; and based on the image generation requirement information and the first reference image, generating (104, 204, 306, 605) a target image using the target image generation manner.