Abstract: An augmentation service receives a request from a client for a set of actions to suggest with respect to content encoded in an image file included with the request. The service sends a first request to a content generation service to obtain the set of actions, including a first prompt that tasks a small language model (SLM) to generate the set of actions based on the content in the image file. The augmentation service replies to the client with at least a portion of the set of actions. The augmentation service receives, from the client, an indication of a selected action of at least the portion of the set of actions and sends a second request to the content generation service to perform the selected action. The second request includes a second prompt that tasks a large language model (LLM) to perform the selected action.

Abstract: An integrated circuit with specialized processing blocks is provided. A specialized processing block may be optimized for machine learning algorithms and may include a multiplier data path that feeds an adder data path. The multiplier data path may be decomposed into multiple partial product generators, multiple compressors, and multiple carry-propagate adders of a first precision. Results from the carry-propagate adders may be added using a floating-point adder of the first precision. Results from the floating-point adder may be optionally cast to a second precision that is higher or more accurate than the first precision. The adder data path may include an adder of the second precision that combines the results from the floating-point adder with zero, with a general-purpose input, or with other dot product terms. Operated in this way, the specialized processing block provides a technical improvement of greatly increasing the functional density for implementing machine learning algorithms.

Abstract: Integrated circuit devices, systems, machine-readable media, and methods are provided to reduce power consumption by supplying a lower voltage to an embedded application-specific integrated circuit (ASIC) block of a programmable logic device than an overall supply voltage while still achieving target performance. An integrated circuit device according this disclosure may include programmable logic circuitry configurable to be set to operate at a first voltage and a first ASIC block configurable to be set to operate at a second voltage different than the first voltage.

Abstract: Integrated circuit devices and circuitry for digital filtering are provided. An integrated circuit device may include a first digital signal processing (DSP) block with first hardened arithmetic circuitry and an output register to store an output of the first DSP block and a second DSP block with second hardened arithmetic circuitry and an input register to receive the output of the first DSP block. An input signal chain may include a first set of registers to provide first input data signals to the first DSP block, a second set of registers to provide second input data signals to the second DSP block, and a third set of registers connected between the first set of registers and the second set of registers to provide delay equal to that of the output register of the first DSP block and the input register of the second DSP block.

Abstract: An integrated circuit includes a digital signal processing circuit. The digital signal processing circuit includes fixed point multiplier circuits, floating point multiplier circuits, and adder circuits. The digital signal processing circuit is configurable to implement a reduction of floating point multiplication operations based on floating point inputs to the digital signal processing circuit using the fixed point multiplier circuits, the floating point multiplier circuits, and the adder circuits.

Abstract: A method includes determining a second torque based on at least a first torque of a steering wheel when there is a collision risk between an intelligent driving device and an obstacle, where the first torque is a torque applied by a user of the intelligent driving device to the steering wheel, a direction of the second torque is opposite to a direction of the first torque, and a value of the second torque minus a value of the first torque is greater than or equal to a first preset threshold; and controlling the steering wheel based on the first torque and the second torque.

Abstract: A neural architecture search (NAS) with a weak predictor comprises: receiving network architecture scoring information; iteratively sampling a search space, wherein the sampling comprises: generating a set of candidate architectures within the search space; learning a first predictor; evaluating performance of the candidate architectures; and based on at least the performance of the set of candidate architectures and the network architecture scoring information, refining the search space to a smaller search space; based on at least the network architecture scoring information, thresholding the performance of candidate architectures to determine scored output candidate architectures; and reporting the scored output candidate architectures. In some examples, the candidate architectures each comprise a machine learning (ML) model, for example a neural network (NN).

Abstract: A mounting structure of a seat for an all terrain vehicle with a vehicle frame and a seat frame, at least one insertion rod is provided at a bottom of the seat frame in a front-and-rear direction, with a rear end of the insertion rod being in snap-fitting engagement with the vehicle frame, the mounting structure comprises a guide carrier fixedly provided on the vehicle frame and having a socket at a rear side of the guide carrier capable of inserting a front end of the insertion rod, with a width of the socket in a left-and-right direction gradually increasing from front to rear, so that the front end of the insertion rod is positioned in an up-and-down direction by snap-fitting engagement into the guide carrier and limited in the left-and-right direction by engagement between inner walls of the guide carrier.

Abstract: The disclosure herein describes training an encoder network to inpaint images with masked portions. A primary encoding process is used to encode a visible portion of a masked input image into encoded token data. The encoded token data is then decoded into both pixel regression output and feature prediction output, wherein both outputs include inpainted image data associated with the masked portion of the masked input image. A pixel regression loss is determined using the pixel regression output and pixel data of an unmasked version of the masked input image. A feature prediction loss is determined using the feature prediction output and ground truth encoding output of the unmasked version of the masked input image. The primary encoding process is then trained using the pixel regression loss and the feature prediction loss, whereby the primary encoding process is trained to encode structural features of input images into encoded token data.

Abstract: A method includes: obtaining, with a sensor of an intelligent driving device, information about a speed limit sign, where the information about the speed limit sign indicates that the speed limit sign is a ramp speed limit sign or an auxiliary road speed limit sign. The method includes determining, based on road information of a first lane on which the intelligent driving device is located and the information about the speed limit sign, whether to alert a user to a speed limit value corresponding to the speed limit sign. The road information of the first lane indicates that the first lane is a main road, a ramp, or an auxiliary road. The method alerts the user to the speed limit value corresponding to the speed limit sign.

Abstract: An internal and external coupling shear resistance locking control system and monitoring method of tunnel surrounding rocks, comprises an internal shear resistance and monitoring system of tunnel surrounding rocks, an external shear resistance and monitoring system of tunnel surrounding rocks, a relative shear displacement monitoring system of the tunnel walls and the roof and floor layers, and an anchor agent installation system; holes are expanded in the anchor cable drilling holes passing through coal-rock interface areas of the coal walls and roof and floor rock layers to form shear misalignment tolerance cavities; monitoring pressure values of measuring bags and liquid discharge in the shear misalignment tolerance cavities to determine shear deformations of the shear misalignment tolerance cavities, monitoring forces on anchor cables, compensating pull rods and shear resistance steel strips, and adjusting the shear resistance of the anchor cable and steel belt through the compensation rod structure.

Abstract: The present invention relates to compounds of formula (I), wherein R1 to R3 and n are as described herein, and their pharmaceutically acceptable salt thereof, and compositions including the compounds and methods of using the compounds.

Abstract: The disclosure herein describes training an encoder network to inpaint images with masked portions. A primary encoding process is used to encode a visible portion of a masked input image into encoded token data. The encoded token data is then decoded into both pixel regression output and feature prediction output, wherein both outputs include inpainted image data associated with the masked portion of the masked input image. A pixel regression loss is determined using the pixel regression output and pixel data of an unmasked version of the masked input image. A feature prediction loss is determined using the feature prediction output and ground truth encoding output of the unmasked version of the masked input image. The primary encoding process is then trained using the pixel regression loss and the feature prediction loss, whereby the primary encoding process is trained to encode structural features of input images into encoded token data.

Abstract: The present shock absorbing structure of an ATV seat for an ATV with a vehicle frame and a seat frame, comprising an elastic support member provided between the seat frame and the vehicle frame, the elastic support member supporting the seat frame and allowing the seat frame to move downward and compress the elastic support member when subjected to an external force. The shock absorbing structure further comprises at least one shock absorber block provided between the seat frame and the vehicle frame, the at least one shock absorber block being either attached to the seat frame and having a gap in an up-and-down direction from the vehicle frame, or attached to the vehicle frame and having a gap in the up-and-down direction from the seat frame. The present shock absorbing structure of an ATV seat provides overall enhanced comfort.

Abstract: The present disclosure describes a digital signal processing (DSP) block that includes a plurality of columns of weight registers and a plurality of inputs configured to receive a first plurality of values and a second plurality of values. The first plurality of values is stored in the plurality of columns of weight registers after being received. Additionally, the DSP block includes a plurality of multipliers configured to simultaneously multiply each value of the first plurality of values by each value of the second plurality of values.

Abstract: An internal and external coupling shear resistance locking control system and monitoring method of tunnel surrounding rocks, comprises an internal shear resistance and monitoring system of tunnel surrounding rocks, an external shear resistance and monitoring system of tunnel surrounding rocks, a relative shear displacement monitoring system of the tunnel walls and the roof and floor layers, and an anchor agent installation system; holes are expanded in the anchor cable drilling holes passing through coal-rock interface areas of the coal walls and roof and floor rock layers to form shear misalignment tolerance cavities; monitoring pressure values of measuring bags and liquid discharge in the shear misalignment tolerance cavities to determine shear deformations of the shear misalignment tolerance cavities, monitoring forces on anchor cables, compensating pull rods and shear resistance steel strips, and adjusting the shear resistance of the anchor cable and steel belt through the compensation rod structure.

Abstract: Example solutions for pluralistic salient object detection are disclosed. A received image shows multiple objects, such as a first object and a second object. A pluralistic object detector is trained to learn tokens. When provided with the image and the first token, it generates a first segmentation mask corresponding to the first object, but not the second object, and when provided with the image and the second token, it generates a second segmentation mask corresponding to at least the second object (and possibly also the first image). When the pluralistic object detector is trained on five tokens, up to five different segmentation masks, each corresponding to a different selection of up to five objects, may be generated. Additionally, a quality predictor is disclosed that assigns quality scores to each of the different segmentation masks, without requiring ground truth for the image.

Abstract: Example solutions for ranking object detection results generate or receive a plurality of segmentation masks each corresponding to one or more images. Each segmentation mask of each plurality of segmentation masks is generated using a different object detector or setting options. A quality predictor assigns a quality score to each segmentation mask, without using ground truth for the image(s). A set (one or more, but less than all) of the highest quality scores is identified for each image. In some examples, an image processing task is performed using the segmentation masks having an assigned quality score that is within the set of highest quality scores. In some examples, only the segmentation mask having the highest quality score for an image is used in the image processing task. In some examples, a quality threshold is provided, and the segmentation masks meeting the quality threshold are used in the image processing task.