Abstract: Systems and methods for the directional interpolation filter for video compression is provided. In some embodiments, the methods and systems for directional interpolation initially determine an angle of directional interpolation for a given pixel in a region of interest. The angle may be determined by performing edge detection for a textural edge in the region of interest and aligning the angle with an angle for the detected edge. Alternatively, the angle may be determined by applying a predictive method. The tap length for the pixel interpolation may also be determined. Determining tap length is an optimization between tap size and desired precision. Once tap length is determined, the systems and methods may interpolate the pixel along the angle by the determined tap length. This interpolation may include interpolation along the angle of integer pixels, or alternatively may include deriving intermediate pixels from integer pixels, and interpolation along the intermediate pixels.

Abstract: Systems and methods for intelligent quantization for video compression are provided. In some embodiments, the methods and systems estimate a rate-distortion function is for video coding. An offset to a reconstruction process is applied, where the offset is determined based upon a frequency dependent variable and a quantization step size. The offset is then derived by transmitting the frequency dependent variable to a decoder. Additionally, a transform coefficient may be estimated as either a Laplacian or Cauchy distributions with a mean of zero and ? as a frequency dependent parameter. The offset at an encoder may be modified based on the frequency dependent variable. Modifying the offset minimizes overall distortion by moving the quantization step size closer to a centroid of the Laplacian or Cauchy distribution. Additionally, a quantization bin size may be altered to optimize distortion, and the system may determine which frequency coefficients to utilize.

Abstract: The present disclosure involves methods, apparatus, and systems for processing comparison in secure multi-party computation. One example method includes, partitioning a first difference (x) between a first share of value a and a first share of value b into N sections, and determining whether a<b based on first indicators and second indicators corresponding to the N sections. The first indicators include a first indicator indicating a comparison result between xj and yj computed based on a Vector Oblivious Shift Evaluation (VOSE) protocol, where xj is a jth section of x, and yj is a jth section of N sections of a second difference (y) between a second secret share of b and a second secret share of a. The second indicators include a second indicator indicating whether xj is a most significant section among sections of the first difference where xj?yj.

Abstract: The methods and systems for model parameter compression includes selecting a total number of bits in which to store a plurality of parameters wherein the number of bits is divided into a signal bit, an exponent set of bits and a fractional set of bits. Additionally, a maximum parameter and a minimum parameter of the plurality of parameters is received. The method maximizes the number of bits in the fractional set of bits subject to two constraints: 1) the fractional set of bits and the exponent set of bits is less than or equal to one less than the total number of bits, and 2) a relational model of the number bits in the exponent set of bits to a range of parameters. Once the fractional set of bits and the exponent set of bits are thus determined, the system may convert the plurality of parameters into a plurality of converted values utilizing a mapping function.

Abstract: Systems and methods for enhanced block prediction for video compression are provided. In some embodiments, the methods and systems for enhanced block prediction initially predict pixels in a video frame to generate predicted blocks. Predicting the pixels includes one or both of an inter prediction and an intra prediction. A weighted combination of the inter prediction and the intra prediction may be used when both prediction methods are employed. After prediction a deep neural network is applied to enhance prediction of the predicted blocks and neighboring reconstructed pixels. The system may determine a number of reconstructed pixels and transmit the number to a decoder. The enhanced prediction may be subtracted from the video frame and then further processed. A determination may be made if blocking artifacts are present, and the system may filter boundaries of the predicted pixels and the neighboring reconstructed pixels.

Abstract: A server obtains a prompt from a user device. The server determines, using an intent matching engine, that the prompt does not match any intent from a predetermined set of intents. The server determines, using a natural language processing engine based on the prompt not matching any intent from the predetermined set of intents, an intent corresponding to the prompt.

Abstract: The invention relates to a downhole system incorporating an electromagnetic disconnect (EMD). The EMD features an actuator rod, a latch mechanism, and a magnetic solenoid, enabling controlled and selective electrical and/or mechanical disconnect of downhole tools. When activated by an electric current, the solenoid triggers the retraction of spring-loaded components, allowing the actuator rod to disengage from the latch. This allows the latch fingers to contract, and when coupled with an electrical connector interface, will disrupt the electrical continuity within the downhole system. Also, the finger contraction will allow for the mechanical disconnect, releasing the tool from the EMD. The mechanism is reusable, allowing for multiple cycles of engagement and disconnect. The system enhances operational efficiency and safety by providing precise control over tool deployment and retrieval in complex wellbore environments.

Abstract: A housing assembly, a mobile energy storage apparatus, and an energy storage system are provided. The housing assembly includes a first housing, a button structure, and a circuit board. The first housing is provided with a first fixing portion on an inner surface of the first housing. The button structure includes a button body passing through a button via hole of the first housing and a resilient arm. The button body is provided with a button protrusion. The circuit board is positioned at one side of the button structure positioned facing away from the button via hole. The circuit board is provided with a control button positioned facing towards the button structure. One of the circuit board and a fixing end of the resilient arm is provided with a positioning portion, and the other of the circuit board and the fixing end is provided with a fitting portion.

Abstract: A portable energy storage apparatus and an energy storage system are provided. The portable energy storage apparatus includes a housing assembly, an energy storage module, a connecting base, and a connecting member. The housing assembly includes a bottom housing and a top housing. The bottom housing is provided with a connecting protrusion protruding from an inner sidewall surface of the bottom housing. The connecting protrusion defines an accommodating cavity therein and an opening in communication with the accommodating cavity. The energy storage module includes an energy storage unit and a fixing member wrapped around an outer periphery of the energy storage unit. The energy storage unit is fixed inside the bottom housing through the fixing member. The fixing member is provided with a fixing portion that defines a first through hole extending therethrough in the first preset direction. The connecting base is embedded in the accommodating cavity through the opening.

Abstract: Systems and methods are disclosed for identifying an object to track. A tracking application accesses images that are being generated by a camera and identifies different objects within each frame, for example, using machine learning or deep learning techniques to learn inherent features of the objects. Each identified object within the frame is compared with objects within visual content items captured by the device. Some objects that are identified within the visual content items may be associated with specific keywords (e.g., a name associated with a picture). The system receives input, such as an audio command via a microphone (e.g., “Track Brian”), identifying the object to track. Upon receipt of the command, the system compares keywords of each object in the frame that matches an object within the visual content items with the received keyword and determines that the object corresponding to the keyword is to be tracked.

Abstract: The present disclosure involves methods, apparatus, and systems for processing equality testing in secure multi-party computation (MPC). In one aspect, a method includes, generating, by a first party of the secure MPC, a difference between a secret share of a first value and a secret share of a second value. During a first iteration, the difference is partitioned into N sections each including M bits. For each section, a random bit is generated, a group of 2M bits are generated based on the random bit, and a selected bit from the group of 2M bits is sent to a second party of the secure MPC based on oblivious transfer (OT) protocol. The random bits of the N sections are concatenated as an input of a second iteration. The method further includes determining whether the first value equals the second value based on performing a plurality of iterations.

Abstract: The present disclosure involves methods, apparatus, and systems for processing equality testing in secure multi-party computation (MPC). In one aspect, a method includes, generating, by a first party of the secure MPC, a difference between a secret share of a first value and a secret share of a second value. During a first iteration, the difference is partitioned into N sections each including M bits. For each section, a random integer is generated, a group of 2M integers are generated based on the random integer, and a selected interger from the group of 2M integers is sent to a second party of the secure MPC based on oblivious transfer protocol. The random integers of the N sections are added as an input of a second iteration. The method further includes determining whether the first value equals the second value based on results of a plurality of iterations.

Abstract: The present disclosure involves methods, apparatus, and systems for processing equality testing in secure multi-party computation (MPC). In one aspect, a method includes, generating, by a first party of the secure MPC, a difference between a secret share of a first value and a secret share of a second value. During a first iteration, the difference is partitioned into N sections each including M bits. For each section, a vector is generated based on Vector Oblivious Shift Evaluation (VOSE) protocol, a first vector comprising 2M bits, and a first secret share of a first indicator is generated. The first indicator indicates whether the section equals a corresponding section of a second difference between a second secret share of the second value and a second secret share of the first value. The first secret share of the first indicator is an arithmetic share.

Abstract: The present disclosure involves methods, apparatus, and systems for processing comparison in secure multi-party computation (MPC). One example method includes, partitioning, by a first party, a first difference (x) between a share of value a and a share of value b into N sections. For each section, the first party generates a first share of a first indicator indicating whether the section is smaller than a corresponding section of a second difference (y) between a second share of b and a second share of a; sends, based on oblivious transfer (OT) protocol, a second share of the first indicator to the second party; generates, based on a second indicator, a first share of a third indicator indicating whether the section is a most significant section that is not equal to the corresponding section of y. The method includes determining whether a<b based on the first indicator and the third indicator.

Abstract: The present application discloses an undergarment including: an undergarment body made of elastic fabric; a zipper; side support panels configured to provide lateral lifting force; and positioning stays configured to provide lower periphery constraints. The undergarment body is provided with two cup portions. The zipper is disposed on the undergarment body between the two cup portions, which when unfastened, opens the undergarment body. The side support panels are disposed on the undergarment body at outer peripheries of the cup portions. The positioning stays are arranged on the undergarment body below the cup portions. This undergarment achieves visual slimming of larger bust profiles while maintaining natural contour aesthetics, exhibiting dual applicability for both athletic and daily wear.

Abstract: A server obtains a prompt from a user device. The server determines a score representing a likelihood that the prompt matches an intent. The server selects, based on the score, a modality to generate a response to the prompt. The modality is selected from a group comprising a workflow for the intent and a generative pretrained transformer engine. The server generates the response using the modality. The server transmits the response to the user device.