Implementations of the present specification disclose a data identification method, apparatus, device, and a computer-readable medium. A solution includes: obtaining a first data set, data samples in the first data set being at least a part of data of a to-be-identified field; obtaining a state transition matrix set generated based on statistics of data samples in a second data set, a data type of the data samples in the second data set being known; determining sample state transition probabilities corresponding to the data samples in the first data set based on the state transition matrix set; determining a ratio between a number of data samples in the first data set whose sample state transition probabilities are greater than a first threshold and a total number of the data samples in the first data set; and determining data corresponding to the to-be-identified field as being of a same data type as the data samples in the second data set in response to that the ratio is greater than a second threshold.

Implementations of the present specification disclose a data identification method, apparatus, device, and a computer-readable medium. A solution includes: obtaining a first data set, data samples in the first data set being at least a part of data of a to-be-identified field; obtaining a state transition matrix set generated based on statistics of data samples in a second data set, a data type of the data samples in the second data set being known; determining sample state transition probabilities corresponding to the data samples in the first data set based on the state transition matrix set; determining a ratio between a number of data samples in the first data set whose sample state transition probabilities are greater than a first threshold and a total number of the data samples in the first data set; and determining data corresponding to the to-be-identified field as being of a same data type as the data samples in the second data set in response to that the ratio is greater than a second threshold.

A machine learning hardware accelerator architecture and associated techniques are disclosed. The architecture features multiple memory banks of very wide SRAM that may be concurrently accessed by a large number of parallel operational units. Each operational unit supports an instruction set specific to machine learning, including optimizations for performing tensor operations and convolutions. Optimized addressing, an optimized shift reader and variations on a multicast network that permutes and copies data and associates with an operational unit that support those operations are also disclosed.

An apparatus for applying dynamic quantization of a neural network is described herein. The apparatus includes a scaling unit and a quantizing unit. The scaling unit is to calculate an initial desired scale factors of a plurality of inputs, weights and a bias and apply the input scale factor to a summation node. Also, the scaling unit is to determine a scale factor for a multiplication node based on the desired scale factors of the inputs and select a scale factor for an activation function and an output node. The quantizing unit is to dynamically requantize the neural network by traversing a graph of the neural network.

A machine learning hardware accelerator architecture and associated techniques are disclosed. The architecture features multiple memory banks of very wide SRAM that may be concurrently accessed by a large number of parallel operational units. Each operational unit supports an instruction set specific to machine learning, including optimizations for performing tensor operations and convolutions. Optimized addressing, an optimized shift reader and variations on a multicast network that permutes and copies data and associates with an operational unit that support those operations are also disclosed.

The present invention provides a target tracking method, system, device and storage medium, which includes: Determining a target area based on the current frame of a training sample, extracting and fusing histogram of oriented gradient (HOG), color naming (CN), and color space HSV features of the target area to obtain a target template; Determining a target function according to the target template and a spatial regularization weight factor; Introducing the Sherman-Morrison formula into the alternating direction method of multipliers (ADMM) to accelerate the solution of the target function and obtain the response value; Iterating the target tracking model when the response value meets the preset confidence threshold until training is completed to obtain a trained target tracking model, and tracking the target in the video to be observed by using the trained target tracking model.

An electronic device: displays an electronic form with a plurality of fields; detects an autofill input that corresponds to a field of the plurality of fields in the electronic form; and in response to detecting the autofill input, updates the electronic form to display fields that have been populated based on a user profile. If the autofill input is associated with a first category of information in the user profile, updating the electronic form includes populating at least two of the plurality of fields using information from the user profile that corresponds to the first category of information. If the autofill input is associated with a second category of information in the user profile, updating the electronic form includes populating at least two of the plurality of fields using information from the user profile that corresponds to the second category of information.

Methods for evaluating sensor data to predict when the sensor should be recalibrated are described. The methods include a model that utilizes current wellbore data as input for the recalibration prediction.

A simulation management service receives a request to perform reinforcement learning for a robotic device. The request can include computer-executable code defining a reinforcement function for training a reinforcement learning model for the robotic device. In response to the request, the simulation management service generates a simulation environment and injects the computer-executable code into a simulation application for the robotic device. Using the simulation application and the computer-executable code, the simulation management service performs the reinforcement learning within the simulation environment.

A machine learning hardware accelerator architecture and associated techniques are disclosed. The architecture features multiple memory banks of very wide SRAM that may be concurrently accessed by a large number of parallel operational units. Each operational unit supports an instruction set specific to machine learning, including optimizations for performing tensor operations and convolutions. Optimized addressing, an optimized shift reader and variations on a multicast network that permutes and copies data and associates with an operational unit that support those operations are also disclosed.