292results about How to "Fast decoding" patented technology

The invention relates to a method for generating a two-dimensional code and implementing mobile payment by a mobile phone. A user can input various payment card information (such as bank card account numbers, credit card account numbers, prepaid card account numbers, and third party payment user account numbers) into the mobile phone, the information is enciphered and stored, and then the two-dimensional code is generated and displayed on a screen of the mobile phone. A merchant scans a pattern of the two-dimensional code through a two-dimensional code identifying and reading device, and after user authentication information (such as two-dimensional code generation time, mobile phone fingerprints and mobile phone positions) and the payment card information are decoded, the user authentication information, the payment card information, consumption amount information of the user, merchant account number information and the like are enciphered and sent to a corresponding payment and settlement system. After receiving the information, a payment and settlement server confirms the payment card information to be true or false according to the two-dimensional code authentication information, and carries out payment transaction. The method has the main advantage that the user can save the information of various payment cards, consumption cards and the like in the mobile phone and realize no-card electronic payment transaction instead of the various payment cards.

A CABAC decoding system and a method are disclosed. The system includes at least a decoding unit group. Each decoding unit group includes N decoding units connected with each other. The Mth decoding unit receives parameter information for decoding bins and bit streams to be decoded, decodes the bins of the bit streams to be decoded, obtains the decoding result of the current decoding unit bin, and sends the updated parameter information to the (M+1)th decoding unit and an output unit. The present invention achieves high decoding rate, and keeps a reasonable cost of hardware resource. So, the present invention provides a high efficient and reasonable decoding solution.

The invention discloses an arithmetic unit used for context arithmetic encoding and decoding. The arithmetic unit comprises a context-based adaptive binary arithmetic coding (CABAC) decoding control module for controlling overall CABAC decoding, a syntactic element decoding control module for controlling the decoding of various syntactic elements, a probability model cache module for caching a probability model required by current decoding, a memory access control module for controlling access to a random access memory (RAM), controlling data interaction between the RAM and the probability model cache module, , saving the data in the probability model cache module and updating the data in the probability model cache module when the probability model cache model contains no probability model required by current decoding, and a decoding engine module for performing arithmetic decoding and calculating bit number that is read in from a code stream in one bin decoding process. The arithmetic unit provided by the invention has high decoding speed, and the consumption of the memory is reduced.

The invention relates to a face recognize system, comprising a network interface module, a dual-core chip SOC module connected with the network interface module, a CPLD address code, a DDR2 high-speed operation memory, and an external program memory, or even with a hard disk with an IDE. The dual-core chip module is connected with the high-speed operation memory via a data bus. The CPLD address code via an EMIF bus is connected with the external program memory NAND Flash or connected with the IDE hard disc too. And the whole system is powered by a power module. The recognize method comprises video code compression, pretreatment, motion check, face positioning, organ positioning, normalizing, character extraction, face comparison, and data communication. The invention has the advantages in high recognize rate and high recognize speed, strong data processing ability, better compatibility with prior detect network, stable operation, easy upgrade and maintenance, with low cost.

In a decoder having an improved LLR (log-likelihood-ratio) update method is provided. The method comprising the steps of: providing a parity check matrix; and using merely a set of parameters on a row of the parity check matrix instead of data of the whole non-zero elements of the parity check matrix free from at least one shifting action after each row updating; thereby saving memory space and process time.

The invention discloses a call speech quality inspection method and device, computer equipment and a storage medium, and is applied to the field of artificial intelligence. The method comprises the following steps: acquiring the call recording data in the quality inspection cycle, wherein the call recording data carries the seat person identifier; using the target speech static decoding network toperform text translation of the call recording data so as to obtain the call text data; performing sensitive word verification on the call text data, extracting the target sensitive words and obtaining the first quality inspection score value based on the target sensitive words; using a multi-dimensional quality inspection scoring model corresponding to the quality inspection scoring table to perform quality inspection scoring on the call recording data and the call text data so as to obtain the second quality inspection score value; and obtaining the comprehensive quality inspection result corresponding to the seat person identifier according to the first quality inspection score value and the second quality inspection score value. The method has low cost and high efficiency in call speech quality inspection and has more objective quality inspection result.

A communication system, network interface and communication port is provided that includes a media local bus. The local bus is connected between a controller and one or more multimedia devices located within a node of the communication system. The controller periodically broadcasts sync signals to the source device, or devices, to synchronize data transmission partitioned into time slots. Each time slot represents is dedicated to a particular data type. Thus, time slot 1 can accommodate packetized data, time slot 2 can accommodate synchronous data, time slot 3 can accommodate control data, and time slot 4 can accommodate isochronous data. Various combinations and variations of those time slots can occur where fewer than four data types can be present within a frame or all four data types can be present. The local bus includes a signal line and one or more data lines.

A connection controlling method of integrating a time synchronous code division multiple access with a automatic-organizing network belongs to the mobile communication field of the communication technology, based on keeping the unitive air interface of TD-SCDMA honeycomb network physic frame invariable construction, the method is a multi-channel synchronization connecting control method based on the preset pattern and distributing pattern control, the time span resource is divided into the control channel and the data channel by the OVSF (Orthogonal Variable Spreading Factor), thus improving the time span reusability through terminal to terminal status to expose the terminal problem, the method maintains the time span status of node within two jumps time span, and realizes the conflict-free wireless connection within two jumps time span; thereby solving the problems of realizing the conflict-free visit time span and performing the reliable alternation communication under integrating network and multi-communication modes.

A decoding system includes a timing loop for use in error recovery mode operations, in which “fast decode” bit values that are used for timing recovery purposes in the normal modes of operation are replaced, at an appropriate iteration in the error recovery mode, with detected bit values from an iterative detector. The replacement occurs after a reread operation in which the timing loop maintains lock.

The invention discloses a method for decoding an image in parallel by a multi-core processor, which comprises the following steps of: sequentially and respectively dividing entropy decoding, rebuilding and loop filtering of each frame image into a plurality of tasks according to macro block rows and putting the tasks into three corresponding queues; and distributing an entropy decoding task, a rebuilding task and a loop filtering task to the multi-core processor for stream processing according to the processes of entropy decoding, rebuilding and loop filtering, and distributing a plurality of cores when the entropy decoding task is completed in the plurality of rebuilding tasks and executing the rebuilding tasks in parallel. Meanwhile, the invention also discloses a device for decoding the image in parallel by the multi-core processor. The multi-core processor can carry out stream processing in combination with the parallel operation during the decoding of the image, and the tasks of each core realize dynamic balance and low delay of decoding, thereby the high-definition images are decoded quickly in real time, and the processing capability of each core is utilized to the maximum limitation.

A MIMO wireless system with Diversity processing is provided having Turbo Codes Decoders for computing orthogonal multipath signals from multiple separate antennas. The invention decodes multipath signals that have arrived at the terminal via different routes after being reflected from buildings, trees or hills. The Turbo Codes Decoder with Diversity processing increases the signal to noise ratio (SNR) more than 6 dB which enables the Wireless system to deliver data rates from up to 600 Mbit / s. A Turbo Codes Decoder block is provided to compute baseband signals from multiple different receiver paths. Several pipelined max-Log-MAP decoders are used for iterative decoding of received data. A Sliding Window of Block N data is used for pipeline operations.

An encoder is provided with an additional coding feature that comprises a time-related term added to a traditional cost function that uses only distortion and byte usage for calculating cost. The time-related term comprises the time that a real decoder needs for decoding a block, and a coefficient. The use of the time-related term will often result in a decision to select a compression mode that is faster to decode than a mode obtained with the traditional cost function. Preferably each receiving terminal belongs to a certain capacity group having specific additional coding features. A single original video is encoded individually for each capacity group according to the additional features of the capacity group in question. The individual encoding relating to each group guarantees that decoding times of frames absolutely or in average remain below the time that real receiving terminals need for decoding encoded frames.

Methods and apparatus for efficient two-stage paging wireless communications systems are described. Wireless terminals are assigned to paging groups. A few first paging message information bits are modulated (using non-coherent modulation) into a first paging signal and communicated from a base station to wireless terminals. WTs wake-up, receive the first paging signal and quickly ascertain whether its paging group should expect a second paging signal, if so, the WT is operated to receive the second paging signal; otherwise, the WT goes back to sleep conserving power. The base station modulates (using coherent modulation) a number of second message information bits into a second paging signal and transmits the signal to WTs. From the information in first and second paging signals, a WT can determine that it is the paged WT and process the paging instructions. The intended paged WT can transmit an acknowledgement signal on a dedicated uplink resource.

The invention discloses a method for realizing setting of a navigating destination through bluetooth push-to-talk. The method is characterized by comprising the following steps of: firstly, starting a push-to-talk device which is connected with a bluetooth module; after matching the bluetooth module with a user mobile phone, dialing a background system phone number by the user mobile phone; after answering the call and understanding a user requirement by a telephone operator of the background system, operating the background system and encoding through DTMF / FSK (Dual Tone Multiple Frequency / Frequency Shift Keying) coder and decoder; returning voice signal data corresponding to map coordinate data to the user mobile phone through a mobile phone network; after receiving the voice signal data by the user mobile phone, sending the voice signal data to the DTMF / FSK decoder in the manner of mobile phone bluetooth; after decoding by the DTMF / FSK decoder, storing in a vehicle master controller MCU (Micro Controller Unit); and lastly, extracting map coordinate information from a memory by a navigating system in a vehicle-mounted terminal and starting a navigating service. The method provided by the invention has the advantages of low cost, high decoding speed and excellent popularizing value.

The embodiment of the invention discloses an MIMO decoding method and device based on deep learning and a storage medium, and the method comprises the steps: building a training data set of MIMO decoding, and enabling the training data set to comprise a plurality of pieces of training data; Training the neural network based on the training data set to obtain a trained neural network model; And when a to-be-decoded MIMO signal is received, inputting the to-be-decoded MIMO signal into the neural network model for MIMO decoding, and then obtaining an MIMO decoding result output by the neural network model. Implementation by the invention, A neural network model used for joint MIMO detection and channel decoding is designed based on deep learning. According to the method, MIMO detection and channel decoding are taken as a joint decoding process, the approximation of an output result of a neural network model is improved through training, the overall performance of MIMO decoding is ensured,and the method has higher decoding accuracy and faster decoding speed.

The invention requests to protect a Chinese speech recognition method combining Transformer and CNN-DFSMN-CTC. The method comprises the steps of: S1, preprocessing a speech signal, and extracting 80-dimensional log mel Fbank features; S2, carrying out convolution on the extracted 80-dimensional Fbank features by using a CNN convolution network; S3, inputting the features into a DFSMN network structure; S4, taking CTC loss as a loss function of an acoustic model, predicting by adopting a Beam search algorithm, and optimizing by using an Adam optimizer; S5, introducing a strong language model Transformer for iterative training until an optimal model structure is achieved; and S6, combining the Transformer with the acoustic model CNN-DFSMN-CTC to carry out adaptation, and carrying out verification on multiple data sets to finally obtain an optimal identification result. According to the method, the recognition accuracy is higher, the decoding speed is higher, the character error rate reaches 11.8% after verification on a plurality of data sets, and the best character error rate reaches 7.8% on an Aidatang data set.

Abstract of Disclosure Turbo Codes Decoder with Diversity processing computes signals from two separate antennas, and by making full use of so-called multipath signals that have arrived at the terminal via different routes after being reflected from buildings, trees or hills. Turbo Codes Decoder with Diversity processing can increase the signal to noise ratio (SNR) more than 6dB which enables 3G system to deliver the data rates from up to 2 Mbit / s. The invention encompasses several improved Turbo Codes Decoder method and apparatus to provide a more suitable, practical and simpler method for implementation a Turbo Codes Decoder in ASIC or DSP codes. (1) Two parallel Turbo Codes Decoder blocks to compute soft-decoded data RXDa, RXDb from two different receiver path. (2) Two pipelined Log-MAP decoders are used for iterative decoding of received data. (3) A Sliding Window of Block N data are used on the input Memory for pipeline operations. (4) The output block N data from the first decoder A are stored in the RAM memory A, and the second decoder B stores output data in the RAM memory B, such that in pipeline mode Decoder A decodes block N data from the RAM memory B while the Decoder B decodes block N data from the RAM memory A at the same clock cycle. (5) Log-MAP decoders are simpler to implement in ASIC and DSP codes with only Adder circuits, and are low-power consumption. (6) Pipelined Log-MAP decoders architecture provides high speed data throughput, one output per clock cycle.

The present invention provides a distortion image correction based two-dimensional code decoding method and a system. The method comprises: acquiring a frame of image that contains a two-dimensional code, and judging the type of the two-dimensional code by determining a plurality of feature points of the two-dimensional code; calling a corresponding matching template according to the type of the two-dimensional code; performing correction by using perspective transformation first, wherein the computation amount of the perspective transformation is small and a quadrilateral pattern with relatively low distortion degree can be realized; filling the matching template with a gray-scale value of each pixel point, wherein rapid decoding is realized by the matching template; if the decoding is failed, then performing correction by starting a curved surface correction algorithm that is relatively complicated and more accurate in correction precision, wherein the curved surface correction can correct a pattern with relatively high distortion degree; performing the perspective transformation on the two-dimensional code pattern subjected to the curved surface correction again for correction. After the curved surface correction, a pattern approximate to a regular quadrangle is obtained and the perspective transformation is performed again, so that the success rate of decoding can be greatly increased.

The invention belongs to the technical field of laser information, and discloses a time control pulse interval laser encoding and decoding method. An encoding end is timed by a Beidou timing module; absolute precise time is obtained; absolute precise integral second moments are taken as references, a demanded quantity of pulse time intervals at a current moment are generated once according to an encoding function; encoding pulse signals are generated at pre-generated time intervals by a timer; a pulse width is controlled; a decoding end obtains the absolute precise moment through the Beidou timing module; the absolute precise moment is synchronous with that of the encoding end; pulse receiving moments are generated according to an encoding function the same to the encoding function of the encoding end; and pulse matriculate wave gates are precisely controlled at the pulse receiving moments by the timer; therefore, the decoding is realized. The method provided by the invention is mainly featured by that the encoding pulse time intervals are different from one another and have no any rule and cycle; the quasi-random distribution encoding has relatively strong anti-decoding capability; the design is simple and is easy to realize; and the method is applicable to fields such as laser semi-active guidance, laser communication and laser radars.

A “STAC Codec” provides lossless audio compression and decompression by processing an audio signal using integer-reversible modulated lapped transforms (MLT) to produce transform coefficients. Transform coefficients are then encoded using a backward-adaptive run-length Golomb-Rice (RLGR) encoder to produce losslessly compressed audio signals. In additional embodiments, further compression gains are achieved via an inter-block spectral estimation and data sorting strategy. Further, compression in the transform domain allows the bitstream to be partially decoded, using the corresponding RLGR decoder, to reconstruct the frequency-domain coefficients. These frequency-domain coefficients are then directly used to speed up various transform-domain based applications such as transcoding media to lossy or other formats, search, identification, visualization, watermarking, etc. In other embodiments, near-lossless compression is achieved by right-shifting transform coefficients by some number of bits such that quantization errors are not perceived as distortion in the decoded audio signal.

The invention discloses an onidirectional multiple line laser bar code reading device and a method thereof. The apparatus comprises a processor, a DAC (digital to analog converter) output circuit, a motor control circuit, a laser, a motor, a laser rotation mechanism, a laser receiver, a front end signal conditioning circuit, a width sampling circuit and a decoding circuit. The processor is connected with the DAC output circuit, the motor control circuit, the decoding circuit and a communication circuit respectively. The DAC output circuit is connected with the laser. The motor control circuit is connected with the motor. The motor drives the laser rotation mechanism to rotate. The laser receiver, the front end signal conditioning circuit, the width sampling circuit and the decoding circuit are successively connected with each other. The method is characterized by: if bar code information is incomplete, reading the bar code information obtained by another scanning line; and then integrating and recombining the back and forth two groups of the information so as to obtain a complete bar code scanning result. Aiming at wrinkling, damage and the reflective barcode, by using the device and the method of the invention, rapid decoding can be realized; a user can feel smooth barcode reading; working efficiency can be increased.

The invention discloses a voice self-registration method, a device, a computer device and a storage medium, which are applied in the field of artificial intelligence. The method comprises the following steps of: a voice registration request is received, wherein the voice registration request carries a registration mark; at least two guide recordings are sent to that client end according to the recording play sequence to enable the client end to play the guiding recording in sequence; registered voice data collected based on each bootstrap recording is recoreded; a target speech static decode network is used for text translation of register speech data to obtain registered text data; Text-to-speech conversion tool is used to convert the registered text data into feedback speech data, and the feedback speech data is played through the client; the confirmation voice data is acquired, and if the confirmation voice data carries the confirmation mark, the user identity registration based onthe registration text data is completed. The method can realize the process of voice self-registration and improve the flexibility and operation convenience of identity registration.

A fast decoding technique for decoding a position of a bit in a pattern provided on a media surface that can generate large amounts of solution candidates quickly by switching or flipping bits and utilizing a recursion scheme. The fast decoding technique may be employed to simultaneously decode multiple dimensions of a pattern on the media surface.