Abstract: A method and an apparatus for detecting a road lane are provided. The method includes acquiring a current road image of a road around a vehicle and inputting the current road image into a deep learning model and detecting a road lane region in the current road image based on a result outputted from the deep learning model. The deep learning model includes a first model device and a second model device. The first model device includes at least one first model subdevice which includes a convolutional neural network and a first recurrent neural network, and the second model device includes at least one second model subdevice which includes a deconvolution neural network and a second recurrent neural network.

Abstract: Fully automatic true triaxial tunnel and underground project model test system, including a triaxial loading device for loading model test piece, automatic data collection and analysis device, power system and control system; triaxial loading device includes test bench, vertical loading system, horizontal front and back, and left and right loading systems, and the vertical, horizontal front and back, and left and right loading systems apply three-way pressure to model test body; test bench functions for supporting, fixing, and providing counter-force; automatic data collection and analysis device includes micro optical fiber sensor embedded in model test piece, optical fiber monitoring system, micro pressure box and strain brick, and can collect multi-field information.

Abstract: A system and method for a computer system for the secure storage, transmittance and access of genetic data includes a coordinator server including a coordinator program arranged to update secure access information, the coordinator server being in communication with a genetic data sequencing server, a genetic data analysing server, and a genetic data storage server, whereby the coordinator server communicates the secure access information in a manner to allow the genetic data storage server to act as a proxy server between the genetic data sequencing server and the genetic data analysing server.

Abstract: Various methods, systems, and processes for mitigating fragmentation in synthetic full backups are disclosed. One or more storage units out of multiple storage units are identified. The multiple storage units include one or more new storage units or one or more existing storage units. The multiple storage units are accessed and a determination is made as to whether the one or more storage units out of the multiple storage units meet a threshold. The threshold is a measure of a characteristic of those one or more storage units. If one or more storage units meet the threshold, those one or more storage units are included in a backup stream, and the backup stream is sent to a backup server.

Abstract: A method and an apparatus for intelligent terrain identification, a vehicle-mounted terminal, and a vehicle are provided. The method includes acquiring an image of a preset driving range in a front driving region of a vehicle; extracting a feature of a road surface from the image, and determining a type of the road surface based on the extracted feature of the road surface, to cause the vehicle to select a corresponding driving strategy according to the type of the road surface. With the method for intelligent terrain identification, the vehicle can identify the type of the current road surface intelligently, and switch the driving strategy of the vehicle automatically according to the type of the road surface, thereby greatly improving convenience of an adaptation function of the vehicle to a terrain, improving driving experience, and reducing probability of safety risks.

Abstract: A method and an apparatus for detecting a road lane are provided. The method includes acquiring a current road image of a road around a vehicle and inputting the current road image into a deep learning model and detecting a road lane region in the current road image based on a result outputted from the deep learning model. The deep learning model includes a first model device and a second model device. The first model device includes at least one first model subdevice which includes a convolutional neural network and a first recurrent neural network, and the second model device includes at least one second model subdevice which includes a deconvolution neural network and a second recurrent neural network.

Abstract: Various methods, systems, and processes for mitigating fragmentation in synthetic full backups are disclosed. One or more storage units out of multiple storage units are identified. The multiple storage units include one or more new storage units or one or more existing storage units. The multiple storage units are accessed and a determination is made as to whether the one or more storage units out of the multiple storage units meet a threshold. The threshold is a measure of a characteristic of those one or more storage units. If one or more storage units meet the threshold, those one or more storage units are included in a backup stream, and the backup stream is sent to a backup server.

Abstract: An adaptive charger can include: a power converter configured to receive an input current from an external power supply, and to generate an output current as a charging current to a load; a current feedback loop configured to compare a first detection signal that represents the input current against a first current reference signal, and to generate a first error signal, where the power converter is configured to regulate the input current according to the first error signal; and the current feedback loop being configured to determine an overload state of the external power supply according to an input voltage of the power converter, where the charger is configured to enter a current limit state when the external power supply is determined to be in the overload state, and where the first current reference signal is gradually reduced until the external power supply recovers to a non-overloaded state.

Abstract: A next generation sequencing (NGS) based method includes applying, for one or more genetic loci, respective NGS data for genotype of a first subject, genotype of a second subject, and genotype of an alleged offspring of the first and second subjects to a statistical model calculating a value representing a likelihood the offspring is a true offspring of the first and second subjects. The NGS data includes genotype and sequencing read of the first tested subject; genotype and sequencing read of the second tested subject; and genotype and sequencing read of the alleged offspring. The statistical model utilizes a probability of the genotype of the first tested subject in a subject population; a probability of the genotype of the second tested subject in a subject population; and a probability of the genotype of the alleged offspring in a subject population.

Abstract: An adaptive charger can include: a power converter configured to receive an input current from an external power supply, and to generate an output current as a charging current to a load; a current feedback loop configured to compare a first detection signal that represents the input current against a first current reference signal, and to generate a first error signal, where the power converter is configured to regulate the input current according to the first error signal; and the current feedback loop being configured to determine an overload state of the external power supply according to an input voltage of the power converter, where the charger is configured to enter a current limit state when the external power supply is determined to be in the overload state, and where the first current reference signal is gradually reduced until the external power supply recovers to a non-overloaded state.

Abstract: The disclosed computer-implemented method for healing images in deduplication storage may include (i) detecting that a data segment stored within deduplication storage is corrupted by executing an algorithm on the data segment to generate a numerical identifier for the data segment and determining that the numerical identifier does not match a previously stored numerical identifier for the data segment, (ii) checking, during a backup operation, whether a digital fingerprint of a data segment of an image being backed up matches a digital fingerprint of the corrupted data segment that was taken prior to corruption and that is already stored within the deduplication storage, and (iii) appending the data segment from the image being backed up to a container that already contains the corrupted data segment rather than replacing the corrupted data segment. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The disclosed computer-implemented method for healing images in deduplication storage may include (i) detecting that a data segment stored within deduplication storage is corrupted by executing an algorithm on the data segment to generate a numerical identifier for the data segment and determining that the numerical identifier does not match a previously stored numerical identifier for the data segment, (ii) checking, during a backup operation, whether a digital fingerprint of a data segment of an image being backed up matches a digital fingerprint of the corrupted data segment that was taken prior to corruption and that is already stored within the deduplication storage, and (iii) appending the data segment from the image being backed up to a container that already contains the corrupted data segment rather than replacing the corrupted data segment. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The present disclosure provides for implementing a two-level fingerprint caching scheme for a client cache and a server cache. The client cache hit ratio can be improved by pre-populating the client cache with fingerprints that are relevant to the client. Relevant fingerprints include fingerprints used during a recent time period (e.g., fingerprints of segments that are included in the last full backup image and any following incremental backup images created for the client after the last full backup image), and thus are referred to as fingerprints with good temporal locality. Relevant fingerprints also include fingerprints associated with a storage container that has good spatial locality, and thus are referred to as fingerprints with good spatial locality. A pre-set threshold established for the client cache (e.g., threshold Tc) is used to determine whether a storage container (and thus fingerprints associated with the storage container) has good spatial locality.

Abstract: An adaptive charger can include: a power converter configured to receive an input current from an external power supply, and to generate an output current as a charging current to a load; a current feedback loop configured to compare a first detection signal that represents the input current against a first current reference signal, and to generate a first error signal, where the power converter is configured to regulate the input current according to the first error signal; and the current feedback loop being configured to determine an overload state of the external power supply according to an input voltage of the power converter, where the charger is configured to enter a current limit state when the external power supply is determined to be in the overload state, and where the first current reference signal is gradually reduced until the external power supply recovers to a non-overloaded state.

Abstract: In one embodiment, a battery charger can include: (i) a step-up converter configured to generate an output signal by boosting a DC input voltage, where a threshold voltage is greater than the DC input voltage; (ii) a charging control circuit configured to receive the output signal from the step-up converter, and to control charging of a battery; (iii) the charging control circuit being configured to regulate the output signal to maintain a charging current for the battery charging as a trickle current when a battery voltage is less than the threshold voltage; and (iv) the charging control circuit being configured to charge the battery directly by the output signal when the battery voltage is greater than the threshold voltage.

Abstract: Disclosed herein are circuits and methods for limiting a charger input current. In one embodiment, a self-adaptive input power charger for charging a battery, can include: (i) a power stage circuit configured to receive an external input power supply that supplies an input voltage and an input current to the charger; (ii) a comparison circuit configured to generate a comparison result indicating that the input power supply has entered a current-limiting state when the input voltage is less than a first reference voltage; (iii) a current regulation circuit configured to generate a first control signal in response to the comparison result; and (iv) a driving control circuit configured to limit the input current by the first control signal.

Abstract: The present invention relates to reference voltage regulating methods and circuits for a constant current driver. In one embodiment, a method can include: setting a reference voltage circuit matching with a current output channel of a constant current source; setting a first resistor of the reference voltage circuit to follow an ideal equivalent resistor of the current output channel, and maintaining a proportion of the first resistor and the ideal equivalent resistor to be no less than a predetermined value M; setting a first current of the reference voltage circuit to follow an ideal output current of the current output channel, and maintaining a proportion of the first current and the ideal output current to be no less than 1/M; and setting a product of the first current and the first resistor to be a reference voltage of the reference voltage circuit.

Abstract: The present invention relates to device matching in an integrated circuit. In one embodiment, an integrated circuit of matched devices can include: N main-devices to be matched by 4×K sub-devices configured to form K device arrays, where each of the device arrays includes four sub-device groups arrayed symmetrically around a vertical axis and a horizontal axis, where each of the sub-device groups includes N sub-devices arrayed with equal distance along a direction of the vertical axis, where K and N are integers, and N is larger than two; metal lead wires arrayed in parallel and with equal distance, and configured to connect the main-devices; a common connecting wire configured to connect common nodes of the sub-devices together; and where four sub-devices arrayed in the four sub-device groups, and other sub-devices arrayed in other sub-device groups, are coupled to form 4×K sub-devices to match the main-devices.

Abstract: A method for managing deduplication reference data may include (1) identifying multiple of data containers configured to store a plurality of deduplicated data segments that are referenced by multiple data objects within a deduplicated data system, (2) maintaining multiple reference databases including (i) a first reference database corresponding to a first subset of the data containers and (ii) a second reference database corresponding to a second subset of the data containers, the second subset differing from the first subset, (3) determining that a data object references at least one segment within a first data container within the first subset but does not reference any data segment within a second data container within the second subset and (4) updating the first reference database with information specifying that the data object references at least one data segment within at least one data container within the first subset of data containers.

Abstract: Methods and circuits related to power regulation are disclosed. In one embodiment, a power regulator for converting an input electrical signal to an output electrical signal to supply power to a load, can include: (i) a power stage having switching devices and a filter; (ii) a regulation signal generator for the switching devices that includes a feedback circuit and a PWM, the feedback circuit receiving an output signal from the power stage, the PWM receiving an output from the feedback circuit, and generating a PWM control signal; (iii) a constant time generator receiving the PWM control signal and generating a constant time signal based on the PWM control signal duty cycle; and (iv) a logic/driving circuit receiving the PWM control signal and the constant time signal, and controlling operation of the switching devices to modulate the output signal from the power stage, and maintaining a pseudo constant operation frequency.