Cirrus Logic Inc.

Cadence Design Systems, Inc.

Sasken Communication Technologies Limited

Sachin P Ghanekar

Dror E Maydan

Rajendra Datar

Pushkar G Patwardhan

Fei Sun

William A Huffman

Manoj Shridhar Soman

Darin S Petkov

Ananda Sarangaram Tharma Ranga Raja

Anoop Deoras

Vijay Pawar

Ravindra Chaugule

Himanshu Singhal

Ravindra Ramkrishna Chaugule

Prasad Nikam

N D Divyakumar

Amit P Sulakhe

Pavan Shridhar Jalwadi

Ravindra Gogte

Samir S Pathak

Sebastian Gracias

Using a buffer sized according to the size of the filters of a convolutional neural network (CNN), a processor may use a read pointer to generate a two-dimensional virtual matrix of inputs. The number of inputs in each row in the two-dimensional virtual matrix of inputs may match the one-dimensional filter size of the cubic filters. The processor may collapse each of the cubic filters to one-dimensional linear arrays and generate a two-dimensional filter matrix from the one-dimensional linear arrays. The convolution computations for a corresponding layer of the CNN therefore reduce to a single matrix multiplication without any memory movement operations. When the buffer is refreshed using a new input frame, the processor may increment the initial read address of each read pointer by one and increment the final read address by one, circling back to the corresponding initial read address.

Systems and methods of buffering and accessing input data for convolution computations

Embodiments may include receiving an input block of data having one or more rows wherein each row includes one or more elements. Embodiments may further include adjusting the input block of data to generate a two-dimensional sorted block of data and identifying at least one position within the two-dimensional sorted block of data that cannot contain a median value or a desired Nth sorted value. Embodiments may also include sorting the two-dimensional block of data along one or more columns to obtain one or more candidate elements that contain the median value or the desired Nth sorted value. Embodiments may include discarding at least one non-candidate element to generate one or more remaining elements and rearranging the one or more remaining elements such that a number of diagonal elements form a column. Embodiments may also include iteratively repeating some of the above operations until a desired value is identified.

System and method for signal processing using sorting based filtering

A system and method for preventing an application program, which is licensed to a customer to be exclusively executed in a processor based on a certain processor design, from being executed properly in unauthorized processors is provided. The system includes a scrambling module and a recovery module. The scrambling module scrambles a selected portion of the application program using an identifier which identifies the authorized processor design. The recovery module adds an unscrambling program to the application program such that when the program is running in a processor, it retrieves a second identifier from the processor and unscrambles the scrambled portion of the application program using the retrieved second identifier. If the second identifier does not correspond to an authorized processor design, the unscrambling operation will incorrectly unscramble the scrambled portion and the application program will not run properly.

System and method for preventing proper execution of an application program in an unauthorized processor

A microprocessor includes a memory interface to obtain data envelopes of a first length, and control logic to implement an instruction to load an initial data envelope of a stream of data values into a buffer, each data value having a second length shorter than the first length, the stream of data values being disposed across successive data envelopes at the memory interface. Another instruction merges current contents of the buffer and the memory interface such that each invocation loads one of the data values into a first register, and moves at least a remainder of the current contents of the memory interface into the buffer for use in a successive invocation. Another instruction loads a reversed representation of a set of data values obtained via the memory interface into a second register. Another instruction implements an FIR computation including a SIMD operation involving multiple data values of the stream and the reversed representation.

Audio digital signal processor

A method and system for automatic gain control of a speech signal in a communication system are disclosed. The gain of the speech signal can be controlled, based on a calculated gain value. This gain value is calculated on the basis of energy calculation and speech activity identification in the speech signal which is done by means of the encoder. Encoding the gain controlled speech signal for transmission follows the step of gain control.

Method and system for automatic gain control of a speech signal

A method, system and computer program product for computationally efficient estimation of the scale factors of one or more frequency bands in an encoder. These scale factors are dependant on a plurality of variables. One of the variables is approximated according to embodiments of the invention. This reduces the complexity of the estimation of scale factors, especially in digital signal processors.

Method for scale-factor estimation in an audio encoder

A method of performing quantization in an audio encoder includes determining a number of bits available in a frame of encoded audio data. Determinations are also made for the maximum transform coefficient value and a distribution of transform coefficient values across the transform coefficient spectrum being encoded. A an estimate for an initial quantization step value is determined from the number of available bits in the frame, the maximum transform coefficient value, and the distribution of coefficient values across the coefficient spectrum.

Methods for selecting an initial quantization step size in audio encoders and systems using the same

A method of generating comfort noise in a communications includes selecting a packet of data from a stream of packets being received across a communications network, the packet comprising a selected number of data samples. Ones of the data samples from the selected packet are played out in a sequence as comfort noise to fill a dead gap. The sequence is reversed when a randomly selected sample in the packet is reached and corresponding ones of the data samples are transmitted across the network as comfort noise after the order is reversed.

Systems and methods for improving sound quality in computer-network telephony application

A method of detecting frequency errors exceeding a predetermined limit in a sampled signal includes the step of determining a peak amplitude of the signal at a tone frequency for a first frame of samples of the sampled signals using a filter having a first amplitude versus frequency response. A peak amplitude of signal at the tone frequency is determined for a second frame of samples of the sampled signal using a filter having a second amplitude versus frequency response. A ratio between the peak amplitude of the first frame and the peak amplitude of the second frame is calculated and compared against a threshold to detect frequency errors exceeding the predetermined limit. Among other things, this method decouples the frequency error detection problem from the twist factor estimation problem.

Frequency error detection methods and systems using the same

An integrated circuit included a plurality of circuit blocks for selectively performing data processing operations in response to a set of instructions. Circuitry dynamically activates and deactivates selected ones of circuit blocks during the execution of the set of instructions.

Dynamically activating and deactivating selected circuit blocks of a data processing integrated circuit during execution of instructions according to power code bits appended to selected instructions

Growing community of inventors

Pune, India

Sachin P Ghanekar