System and method for false positive example mitigation in pulsed sound detectors

Abstract

A data enhancement method includes receiving audio stream data associated with at least one pulse event, receiving a tag associated with the audio stream data, and detecting at least one peak of the at least one pulse event using a start point detector. The method also includes extracting at least one positive sample of the audio stream data associated with the at least one pulse event. The method also includes applying a tag associated with the audio stream data to the at least one positive sample and extracting at least one negative sample of the audio stream data associated with the at least one pulse event. The method also includes enhancing training data based on the at least one positive sample and the at least one negative sample, and training at least one machine learning model using the enhanced training data.

Description

Technical Field

[0001] The present disclosure relates to computer systems with artificial intelligence capabilities, including neural networks.In an embodiment, the present disclosure relates to mitigating false positive impulse sound detection. Background Art

[0002] The use of machine learning models to detect and identify various impulse sounds (e.g., gunshots, vehicle crashes, etc.) in audio recordings and live acquisitions has several useful applications ranging from physical security to content retrieval. However, successful deployment of such machine learning models can be challenging because different types of impulse sounds share common temporal and spectral characteristics that may be dictated by the surrounding acoustic environment (e.g., rather than the impact event type itself). As a result, the output of such machine learning models may include a relatively high false positive rate, which may significantly reduce the practical usefulness of such systems. Summary of the invention

[0003] One aspect of the disclosed embodiment includes a computer-implemented method for data enhancement. The method includes receiving audio stream data associated with at least one impulse event, receiving a label associated with the audio stream data, and detecting at least one peak of the at least one impulse event using an onset detector algorithm. The method also includes extracting at least one positive sample of the audio stream data associated with the at least one impulse event, wherein the at least one positive sample includes a portion of the audio stream data associated with a first time value of the audio stream data. The method also includes labeling the at least one positive sample using the label associated with the audio stream data to produce at least one labeled positive sample, and extracting at least one negative sample of the audio stream data associated with the at least one impulse event, wherein the at least one negative sample does not include the portion of the audio stream data associated with the first time value of the audio stream data. The method also includes enhancing machine learning training data based on the at least one labeled positive sample and the at least one negative sample.

[0004] Another aspect of the disclosed embodiment includes a system for training a pulse event detection machine learning model. The system includes a processor and a memory. The memory includes instructions that, when executed by the processor, cause the processor to: receive audio stream data associated with at least one pulse event; receive a label associated with the audio stream data; detect at least one peak of at least one pulse event using a starting point detector algorithm; extract at least one positive sample of the audio stream data associated with at least one pulse event, wherein at least one positive sample includes a portion of the audio stream data associated with a first time value of the audio stream data; label at least one positive sample using the label associated with the audio stream data to produce at least one labeled positive sample; extract at least one negative sample of the audio stream data associated with at least one pulse event, wherein at least one negative sample does not include a portion of the audio stream data associated with the first time value of the audio stream data; enhance the machine learning training data based on at least one labeled positive sample and at least one negative sample; and train the pulse event detection machine learning model using the enhanced machine learning training data.

[0005] Another aspect of the disclosed embodiment includes an apparatus for pulse event detection. The apparatus includes a processor and a memory. The memory includes instructions that, when executed by the processor, cause the processor to: receive audio stream data associated with at least one pulse event; receive a label associated with the audio stream data; use a starting point detector algorithm to detect at least one peak of at least one pulse event, wherein the starting point detector is configured to use at least one of an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data to detect at least one peak; extract at least one positive sample of the audio stream data associated with at least one pulse event, wherein the at least one positive sample includes a portion of the audio stream data associated with a first time value of the audio stream data; and use a starting point detector algorithm to detect at least one peak of the at least one pulse event. Labeling at least one positive sample with a label associated with audio stream data to produce at least one labeled positive sample; extracting at least one negative sample of the audio stream data associated with at least one impulse event, wherein the at least one negative sample does not include a portion of the audio stream data associated with a first time value of the audio stream data; based on the at least one labeled positive sample and the at least one negative sample, enhancing machine learning training data using at least one of spectral masking, spectral shifting, at least one background sound, and at least one impulse response associated with an acoustic environment; training at least one machine learning model using the enhanced machine learning training data; and using the at least one machine learning model to detect one or more impulse events associated with audio data provided as input to the at least one machine learning model. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 A system for training a neural network in accordance with the principles of the present disclosure is generally illustrated.

[0007] Figure 2 A computer-implemented method for training and utilizing a neural network in accordance with the principles of the present disclosure is generally illustrated.

[0008] Figure 3A-3C Various pulse events and window selections in accordance with the principles of the present disclosure are generally illustrated.

[0009] Figure 4 is a flow chart generally illustrating a method of data enhancement according to the principles of the present disclosure.

[0010] Figure 5 Depicted is a schematic diagram of the interaction between a computer-controlled machine and a control system in accordance with the principles of the present disclosure.

[0011] Figure 6 Depicted is a device configured to control a vehicle according to the principles of the present disclosure. Figure 5 Schematic diagram of a control system of a vehicle, which can be a partially autonomous vehicle, a fully autonomous vehicle, a partially autonomous robot, or a fully autonomous robot.

[0012] Figure 7 Depicted Figure 5 Schematic diagram of a control system configured to control a manufacturing machine, such as a punch, cutter, or gun drill, of a manufacturing system (such as part of a production line).

[0013] Figure 8 Depicted Figure 5 Schematic diagram of a control system configured to control a power tool, such as a drill or driver, having an at least partially autonomous mode.

[0014] Fig. 9 Depicted Figure 5 Schematic diagram of a control system configured to control an automated personal assistant.

[0015] Fig.10 Depicted Figure 5 Schematic diagram of a control system configured to control a monitoring system, such as a controlled access system or a surveillance system.

[0016] Fig.11 Depicted Figure 5 Schematic diagram of a control system configured to control an imaging system, such as an MRI device, an x-ray imaging device, or an ultrasound device. DETAILED DESCRIPTION

[0017] Embodiments of the present disclosure are described herein. However, it will be understood that the disclosed embodiments are merely examples, and other embodiments may take various forms and alternative forms. The figures are not necessarily drawn to scale; some features may be enlarged or minimized to show the details of a particular component. Therefore, the specific structural and functional details disclosed herein should not be interpreted as restrictive, but only as a representative basis for teaching those skilled in the art to adopt the embodiments in various ways. As will be understood by those of ordinary skill in the art, the various features illustrated and described with reference to any one of the figures may be combined with the features illustrated in one or more other figures to produce embodiments that are not explicitly illustrated or described. The combination of the illustrated features provides representative embodiments for typical applications. However, for specific applications or embodiments, various combinations and modifications of features consistent with the teachings of the present disclosure may be desired.

[0018] As described, the use of machine learning models to detect and identify various impulse sounds (e.g., gunshots, vehicle collisions, etc.) in audio recordings and real-time acquisitions has several useful applications from physical security to content retrieval. Data enhancement for audio event detection is commonly used in scientific literature to increase the robustness and sensitivity of artificial intelligence-based machine learning models. Typically, such systems rely on audio data enhancement techniques, including time offset, frequency offset, noise addition, time-frequency masking, etc. Additionally, such systems typically depend directly on the number, type, and quality of samples provided during the training phase. For audio classification systems, this aspect is even more important because the number of labeled audio samples is typically at least one or two orders of magnitude smaller than other fields (such as images or text). Sound event classification systems are typically pre-trained on a large corpus of weakly labeled audio samples (e.g., without time tags) and then fine-tuned for specific applications. The fine-tuning stage is particularly sensitive to the quality of the input samples.

[0019] However, successful deployment of such machine learning models can be challenging because different types of impulse sounds share common temporal and spectral properties that may be dictated by the surrounding acoustic environment (e.g., rather than the type of impulse event itself). As a result, the output of such machine learning models may include a relatively high false positive rate, which may significantly reduce the practical usefulness of such systems.

[0020] Therefore, systems and methods configured to improve the mitigation of false positive impulse event detection (such as those described herein) may be desirable. In some embodiments, the systems and methods described herein can be configured to provide data augmentation for machine learning training data, which is customized to enhance available recordings of impulse sounds of interest to enhance false positive resilience when using various data-driven techniques (e.g., neural networks, machine learning models, etc.) to identify target sounds.

[0021] In some embodiments, the systems and methods described herein can be configured to train a data-driven technique (e.g., a neural network, a machine learning model, etc.) to classify input sounds into M categories. When training a neural network for sound event detection, the input to the network is typically an audio window of fixed duration. At inference time, the machine learning model is provided with windows of the same length, typically with a hop size much shorter than the window length (e.g., a 2 second window is fed every 0.2 seconds, or any other suitable length). This typically provides the machine learning model with multiple views of the same sound over time.

[0022] The systems and methods described herein can be configured to identify impulsive sounds in audio data provided to a corresponding machine learning model. The systems and methods described herein can be configured to characterize impulsive sounds by two phases: a fast transient phase (e.g., a pulse, which can be referred to as an onset or peak) and a tail phase. The systems and methods described herein can be configured to identify the phase of the type of impulsive sound. For example, Figure 3A As generally illustrated in FIG, a first impact event (e.g., a gunshot or other suitable impact event) (generally illustrated as Impact Event 1) and a second impact event (e.g., glass breaking or other suitable impact event) (generally illustrated as Impact Event 2) may sound very similar if an observer hears only the transient phase, but may sound very different if an observer hears each tail phase of the corresponding impulse sound. Additionally or alternatively, Impact Event 1 and a third impact event (e.g., a door slammed or other suitable impact event) (generally illustrated as Impact Event 3) may sound similar in the corresponding tail phase (e.g., if both are recorded in the same acoustic environment), but may exhibit different transient phases.

[0023] In some embodiments, the systems and methods described herein can be configured to, given a recording containing one or more impulsive sounds of interest, generate a set of positive windows (e.g., samples that can be classified as sounds of interest) and a set of negative windows (e.g., samples that can be classified as not of interest). The systems and methods described herein can be configured to increase the accuracy of sound event detection for impulsive sounds belonging to a particular class while rejecting other impulsive sounds. For example, the systems and methods described herein can be configured to train a gunshot detector that can reject all other types of impulsive sounds (e.g., breaking glass, fireworks, a falling box, a door slamming, etc.).

[0024] The systems and methods described herein can be configured to generate a set of positive windows and a set of negative windows for each available sample containing an impulse sound of interest. The systems and methods described herein can be configured to apply random window extraction (e.g., random time offset) using an enhancement scheme to provide training samples labeled as specific impulse events (e.g., gunshots or other suitable impulse events) that contain only a portion of the sound of interest (e.g., only the tail, only the starting point without much tail, the last part of the tail, silence before the event, a small portion of the pulse, etc.).

[0025] The systems and methods described herein can be configured to generate negative samples whenever the window does not include the pulse of the sound and sufficient background around it. The systems and methods described herein can be configured to learn that a particular pulse sound (e.g., a gunshot) consists of a particular pulse followed by a particular tail, while the tail or pulse alone are not sufficient to characterize the type of pulse sound.

[0026] In some embodiments, the systems and methods described herein may be configured such that, given an audio recording and associated tags, Figure 3A-3C As generally illustrated in FIG. 1 , an onset detector (e.g., which may include an onset detector algorithm executable by a processor such as those described herein or any suitable processor, processing device, and / or computing device) is used to detect peak(s) of(s) impulse events in an input sample (e.g., an audio recording). The systems and methods described herein may be configured to, for a given time period at T p For each identified peak at , a set of positive windows (e.g., WIN 1, WIN 2, and WIN 3) is extracted such that each window contains at least T before the peak. b seconds and at least T after the peak a seconds. Positive samples inherit the labels of the original records. T is selected based on the machine learning model to be trained. b and T aFor example, for a deep neural network with an input of 2 seconds, T b and T a The reference values ​​of are 100 milliseconds and 1.5 seconds respectively. It should be understood that given any suitable input sample, T b and T a can be any suitable value for any suitable machine learning model or neural network.

[0027] The systems and methods described herein can be configured to extract, for each identified peak, a set of negative windows (e.g., WIN4, WIN5, WIN6, and WIN7) such that each window does not meet at least one of the criteria defined for the positive window. Negative samples are associated with different labels (e.g., "other sound events"). It should be understood that the systems and methods described herein can be configured to any suitable number of positive windows and / or negative windows.

[0028] The systems and methods described herein can be configured to collect all positive and negative samples, as well as other samples used for training. The systems and methods described herein can be configured to perform various other data enhancements based on the machine learning model to be trained, such as: spectral masking and spectral shifting (e.g., along frequency); mixing with several background sounds (e.g., wind, traffic, road noise, etc.); etc.

[0029] The systems and methods described herein can be configured to apply impulse responses from real and / or simulated acoustic environments. The systems and methods described herein can be configured to train machine learning models using backpropagation. For example, given a digital audio recording represented by a sequence x[n], a series of windows w of fixed duration N samples are extracted from x with a fixed hop step of H samples. i [n]. The systems and methods described herein may be configured to calculate an energy envelope e[i], which may be defined as:

[0030]

[0031] The systems and methods described herein may be configured to set the energy lower limit e f is defined as the 30% percentile of the energy envelope (or other suitable value). The systems and methods described herein may be configured to convert the energy peak e p Defined as the maximum value of e[i]. The systems and methods described herein can be configured to define a dynamic range α∈(0,1). The systems and methods described herein can be configured to extract the event starting point O n ={i|e[i+1]-e[i]>e p·α}. The systems and methods described herein may be configured to extract event offset O f ={i|e[i+1]-e[i]>e f · 1 / α}. The systems and methods described herein may be configured to group starting points and offsets The peak values ​​are obtained from The starting point of the grouping is obtained.

[0032] like Figure 3C As generally illustrated in , the systems and methods described herein can be configured such that, given a window of fixed duration, any possible window strictly included in the "positive window region" is a positive window, and therefore inherits the original record label. Any window that does not completely fit into the "positive window region" is a negative window.

[0033] In some embodiments, the system and method described herein can be configured to receive audio stream data associated with at least one impulse event.The system and method described herein can be configured to receive a label associated with the audio stream data.The system and method described herein can be configured to use a starting point detector to detect at least one peak value of at least one impulse event.The starting point detector can be configured to use at least one of the following items to detect at least one peak value: an energy envelope associated with the audio stream data, at least one energy lower limit associated with the audio stream data, at least one energy peak associated with the audio stream data, at least one dynamic range associated with the audio stream data, other suitable data or information or a combination thereof.The system and method described herein can be configured to extract at least one positive sample of the audio stream data associated with at least one impulse event.At least one positive sample can include a part of the audio stream data associated with the first time value of the audio stream data.

[0034] The systems and methods described herein may be configured to apply a label associated with the audio stream data to at least one positive sample. The at least one positive sample may extend a first time period before a first time value and extend a second time period after the first time value. The systems and methods described herein may be configured to extract at least one negative sample of the audio stream data associated with at least one impulse event. The at least one negative sample may not include a portion of the audio stream data associated with the first time value of the audio stream data. The at least one negative sample may include a label different from the label associated with the audio stream data. The at least one negative sample and the at least one positive sample may not overlap.

[0035] The systems and methods described herein may be configured to enhance training data based on at least one positive sample and at least one negative sample. The systems and methods described herein may be configured to enhance training data using at least one of the following: spectrum masking, spectrum shift, at least one background sound, at least one impulse response associated with an acoustic environment, other suitable enhancement techniques or combinations thereof. The systems and methods described herein may be configured to train (e.g., using at least one back propagation technique or other suitable techniques) at least one machine learning model using enhanced training data. The systems and methods described herein may be configured to use at least one machine learning model to detect one or more impulse events associated with audio data provided to at least one machine learning model as input. The at least one machine learning model may be configured to at least detect a class of sounds associated with the at least one impulse event.

[0036] Figure 1 A system 100 for training a neural network is shown. The system 100 may include an input interface for accessing training data 102 for the neural network. For example, Figure 1 As shown, the input interface may be constituted by a data storage interface 104, which may access the training data 102 from a data storage device 106. For example, the data storage interface 104 may be a memory interface or a persistent storage interface, such as a hard disk or SSD interface, or a personal area network, a local area network, or a wide area network interface, such as a Bluetooth, Zigbee, or Wi-Fi interface, or an Ethernet or fiber optic interface. The data storage device 106 may be an internal data storage device of the system 100 (such as a hard disk or SSD), but may also be an external data storage device, such as a network accessible data storage device.

[0037] In some embodiments, data storage 106 may also include a data representation 108 of an untrained version of the neural network, which may be accessed by system 100 from data storage 106. However, it will be appreciated that training data 102 and data representation 108 of the untrained neural network may also each be accessed from a different data storage (e.g., via a different subsystem of data storage interface 104). Each subsystem may be of the type described above with respect to data storage interface 104.

[0038] In some embodiments, the data representation 108 of the untrained neural network may be generated internally by the system 100 based on design parameters for the neural network, and thus may not be explicitly stored on the data storage device 106. The system 100 may also include a processor subsystem 110, which may be configured to provide an iterative function to replace a layer stack of the neural network to be trained during operation of the system 100. Here, the corresponding layers of the replaced layer stack may have mutually shared weights and may receive the output of the previous layer as input, or, for the first layer of the layer stack, receive an initial activation and a portion of the input of the layer stack.

[0039] The processor subsystem 110 may also be configured to iteratively train the neural network using the training data 102. Here, the training iterations by the processor subsystem 110 may include a forward propagation portion and a backward propagation portion. The processor subsystem 110 may be configured to perform the forward propagation portion by determining an equilibrium point of the iterative function (at which the iterative function converges to a fixed point) and by providing the equilibrium point as a substitute for the output of the stack of layers in the neural network, in addition to defining other operations that may be performed for the forward propagation portion, wherein determining the equilibrium point includes using a numerical root finding algorithm to find a root solution of the iterative function minus its input.

[0040] The system 100 may also include an output interface for outputting a data representation 112 of the trained neural network, which may also be referred to as trained model data 112. Figure 1 As illustrated, the output interface may be comprised of a data storage interface 104, which in these embodiments is an input / output ("IO") interface, via which trained model data 112 may be stored in a data storage device 106. For example, during or after training, a data representation 108 defining an "untrained" neural network may be at least partially replaced by a data representation 112 of a trained neural network, as the parameters of the neural network (such as weights, hyperparameters, and other types of parameters of the neural network) may be adapted to reflect the training of the training data 102. This is also illustrated by reference numerals 108, 112. Figure 1 , reference numerals 108, 112 refer to the same data record on the data storage device 106. In some embodiments, the data representation 112 may be stored separately from the data representation 108 defining the "untrained" neural network. In some embodiments, the output interface may be separate from the data storage interface 104, but may generally be of the type described above for the data storage interface 104.

[0041] Figure 2A data annotation / enhancement system 200 is depicted to implement a system for annotating and / or enhancing data. The data annotation system 200 may include at least one computing system 202. The computing system 202 may include at least one processor 204, which may be operably connected to a memory unit 208. The processor 204 may include one or more integrated circuits that implement the functionality of a central processing unit (CPU) 206. The CPU 206 may be a commercially available processing unit that implements an instruction set, such as one of the x86, ARM, Power, or MIPS instruction set families.

[0042] During operation, CPU 206 may execute stored program instructions retrieved from memory unit 208. The stored program instructions may include software that controls the operation of CPU 206 to carry out the operations described herein. In some embodiments, processor 204 may be a system on a chip (SoC) that integrates the functionality of CPU 206, memory unit 208, network interfaces, and input / output interfaces into a single integrated device. Computing system 202 may implement an operating system for managing various aspects of operation.

[0043] The memory unit 208 may include volatile memory and non-volatile memory for storing instructions and data. Non-volatile memory may include solid-state memory, such as NAND flash memory, magnetic and optical storage media, or any other suitable data storage device that retains data when the computing system 202 is deactivated or powered off. Volatile memory may include static and dynamic random access memory (RAM) that stores program instructions and data. For example, the memory unit 208 may store a machine learning model 210 or algorithm, a training data set 212 for the machine learning model 210, and an original source data set 216.

[0044] The computing system 202 may include a network interface device 222 configured to provide communications with external systems and devices. For example, the network interface device 222 may include a wired and / or wireless Ethernet interface as defined by the Institute of Electrical and Electronics Engineers (IEEE) 802.11 family of standards. The network interface device 222 may include a cellular communication interface for communicating with a cellular network (e.g., 3G, 4G, 5G). The network interface device 222 may also be configured to provide a communication interface to an external network 224 or a cloud.

[0045] The external network 224 may be referred to as the World Wide Web or the Internet. The external network 224 may establish a standard communication protocol between computing devices. The external network 224 may allow information and data to be easily exchanged between computing devices and the network. One or more servers 230 may communicate with the external network 224.

[0046] Computing system 202 may include input / output (I / O) interface 220, which may be configured to provide digital and / or analog input and output. I / O interface 220 may include an additional serial interface (eg, a universal serial bus (USB) interface) for communicating with external devices.

[0047] The computing system 202 may include a human-machine interface (HMI) device 218, which may include any device that enables the system 200 to receive control inputs. Examples of input devices may include human-machine interface inputs, such as keyboards, mice, touch screens, voice input devices, and other similar devices. The computing system 202 may include a display device 232. The computing system 202 may include hardware and software for outputting graphics and text information to the display device 232. The display device 232 may include an electronic display screen, a projector, a printer, or other suitable devices for displaying information to a user or operator. The computing system 202 may also be configured to allow interaction with a remote HMI and a remote display device via a network interface device 222.

[0048] System 200 may be implemented using one or more computing systems. Although the example depicts a single computing system 202 that implements all of the described features, it is intended that the various features and functions may be separate and implemented by multiple computing units that communicate with each other. The specific system architecture selected may depend on a variety of factors.

[0049] The system 200 may implement a machine learning algorithm 210 that is configured to analyze a raw source data set 216. The raw source data set 216 may include raw or unprocessed sensor data that may represent an input data set for the machine learning system. The raw source data set 216 may include videos, video clips, images, text-based information, and raw or partially processed sensor data (e.g., a radar map of an object). In some embodiments, the machine learning algorithm 210 may be a neural network algorithm that is designed to perform a predetermined function. For example, the neural network algorithm may be configured in an automotive application to identify pedestrians in a video image.

[0050] The computer system 200 may store a training data set 212 for the machine learning algorithm 210. The training data set 212 may represent a set of previously constructed data for training the machine learning algorithm 210. The machine learning algorithm 210 may use the training data set 212 to learn weighting factors associated with the neural network algorithm. The training data set 212 may include a set of source data with corresponding outputs or results that the machine learning algorithm 210 attempts to replicate via the learning process. In this example, the training data set 212 may include source videos with and without pedestrians and corresponding presence and location information. The source video may include various scenes in which pedestrians are identified.

[0051] The machine learning algorithm 210 can operate in a learning mode using the training data set 212 as input. The machine learning algorithm 210 can be executed in multiple iterations using data from the training data set 212. In each iteration, the machine learning algorithm 210 can update the internal weighting factors based on the achieved results. For example, the machine learning algorithm 210 can compare the output results (e.g., annotations) with those included in the training data set 212. Since the training data set 212 includes expected results, the machine learning algorithm 210 can determine when the performance is acceptable. After the machine learning algorithm 210 achieves a predetermined performance level (e.g., 100% consistency with the results associated with the training data set 212), the machine learning algorithm 210 can be executed using data that is not in the training data set 212. The trained machine learning algorithm 210 can be applied to a new data set to generate annotated data.

[0052] The machine learning algorithm 210 may be configured to identify specific features in the raw source data 216. The raw source data 216 may include multiple instances or input data sets for which supplementary results are desired. For example, the machine learning algorithm 210 may be configured to identify the presence of pedestrians in a video image and annotate the presence. The machine learning algorithm 210 may be programmed to process the raw source data 216 to identify the presence of specific features. The machine learning algorithm 210 may be configured to identify the features in the raw source data 216 as predetermined features (e.g., pedestrians). The raw source data 216 may be obtained from a variety of sources. For example, the raw source data 216 may be actual input data collected by the machine learning system. The raw source data 216 may be machine generated for testing the system. As an example, the raw source data 216 may include raw video images from a camera.

[0053] In an example, the machine learning algorithm 210 can process the raw source data 216 and output an indication of an image representation. The output can also include an enhanced representation of the image. The machine learning algorithm 210 can generate a confidence level or factor for each generated output. For example, a confidence value exceeding a predetermined high confidence threshold can indicate that the machine learning algorithm 210 is confident that the identified feature corresponds to a particular feature. A confidence value less than a low confidence threshold can indicate that the machine learning algorithm 210 has some uncertainty about the existence of a particular feature.

[0054] In some embodiments, the system 200 can be configured to train the machine learning model 210 to classify the impulse sound. For example, the system 200 can receive the audio stream data associated with at least one impulse event. The system 200 can receive the label associated with the audio stream data. The system 200 can use the starting point detector to detect at least one peak of at least one impulse event. The starting point detector can be configured to detect at least one peak using at least one of the following: an energy envelope associated with the audio stream data, at least one energy lower limit associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data, other suitable data or information, or a combination thereof.

[0055] The system 200 may extract at least one positive sample of the audio stream data associated with the at least one impulse event. The at least one positive sample may include a portion of the audio stream data associated with a first time value of the audio stream data. The system 200 may apply a label associated with the audio stream data to the at least one positive sample. The at least one positive sample may extend a first time period before the first time value and extend a second time period after the first time value. The system 200 may extract at least one negative sample of the audio stream data associated with the at least one impulse event. The at least one negative sample may not include a portion of the audio stream data associated with the first time value of the audio stream data. The at least one negative sample may include a label different from the label associated with the audio stream data. The at least one negative sample and the at least one positive sample may not overlap.

[0056] The system 200 can enhance the training data based on at least one positive sample and at least one negative sample. The system 200 can enhance the training data using at least one of the following: spectral masking, spectral shifting, at least one background sound, at least one impulse response associated with the acoustic environment, other suitable enhancement techniques, or combinations thereof. The system 200 can use the enhanced training data to train (e.g., using at least one back-propagation technique or other suitable technique) the machine learning model 210. The system 200 can use the machine learning model 210 to detect one or more impulse events associated with audio data provided as input to the machine learning model 210. The machine learning model 210 can be configured to at least detect a class of sounds associated with at least one impulse event. Additionally or alternatively, the machine learning model 210 can be configured to perform any suitable function, such as the functions described herein. Figure 6-Figure 11 Those features described.

[0057] Figure 4 is a flow chart generally illustrating a data enhancement method 400 in accordance with the principles of the present disclosure. At 402, the method 400 receives audio stream data associated with at least one impulse event.

[0058] At 404, method 400 receives tags associated with audio stream data.

[0059] At 406 , method 400 detects at least one peak of at least one pulse event using an onset detector.

[0060] At 408, method 400 extracts at least one positive sample of the audio stream data associated with the at least one impulse event. The at least one positive sample may include a portion of the audio stream data associated with a first time value of the audio stream data.

[0061] At 410 , method 400 applies a label associated with the audio stream data to at least one positive sample.

[0062] At 412, method 400 extracts at least one negative sample of the audio stream data associated with the at least one impulse event. The at least one negative sample may not include a portion of the audio stream data associated with the first time value of the audio stream data.

[0063] At 414 , method 400 augments the training data based on at least one positive sample and at least one negative sample.

[0064] At 416, method 400 trains at least one machine learning model using the augmented training data. Method 400 may include using at least one machine learning model to detect one or more impulse events using audio data provided to the at least one machine learning model.

[0065] Figure 5 A schematic diagram of the interaction between a computer controlled machine 500 and a control system 502 is depicted. The computer controlled machine 500 includes an actuator 504 and a sensor 506. The actuator 504 may include one or more actuators, and the sensor 506 may include one or more sensors. The sensor 506 is configured to sense a condition of the computer controlled machine 500. The sensor 506 may be configured to encode the sensed condition into a sensor signal 508 and transmit the sensor signal 508 to the control system 502. Non-limiting examples of the sensor 506 include video, radar, lidar, ultrasonic, and motion sensors. In some embodiments, the sensor 506 is an optical sensor configured to sense an optical image of an environment close to the computer controlled machine 500.

[0066] The control system 502 is configured to receive sensor signals 508 from the computer controlled machine 500. The control system 502 may also be configured to calculate and transmit actuator control commands 510 to actuators 504 of the computer controlled machine 500 depending on the sensor signals, as explained below.

[0067] like Figure 5 As shown in , the control system 502 includes a receiving unit 512. The receiving unit 512 can be configured to receive the sensor signal 508 from the sensor 506 and transform the sensor signal 508 into an input signal x. In an alternative embodiment, the sensor signal 508 is directly received as the input signal x without the receiving unit 512. Each input signal x can be a part of each sensor signal 508. The receiving unit 512 can be configured to process each sensor signal 508 to generate each input signal x. The input signal x can include data corresponding to the image recorded by the sensor 506.

[0068] The control system 502 includes a classifier 514. The classifier 514 can be configured to classify the input signal x into one or more labels using a machine learning (ML) algorithm, such as a neural network described above. The classifier 514 is configured to be parameterized by parameters, such as the parameters described above (e.g., parameter θ). The parameter θ can be stored in and provided by a non-volatile storage device 516. The classifier 514 is configured to determine an output signal y based on the input signal x. Each output signal y includes information for assigning one or more labels to each input signal x. The classifier 514 can transmit the output signal y to a conversion unit 518. The conversion unit 518 is configured to convert the output signal y into an actuator control command 510. The control system 502 is configured to transmit the actuator control command 510 to the actuator 504, and the actuator 504 is configured to actuate the computer-controlled machine 500 in response to the actuator control command 510. In some embodiments, actuator 504 is configured to actuate computer-controlled machine 500 directly based on output signal y.

[0069] When the actuator control command 510 is received by the actuator 504, the actuator 504 is configured to perform an action corresponding to the associated actuator control command 510. The actuator 504 may include control logic configured to transform the actuator control command 510 into a second actuator control command that is utilized to control the actuator 504. In one or more embodiments, the actuator control command 510 may be utilized to control a display instead of or in addition to the actuator.

[0070] In some embodiments, the control system 502 includes the sensor 506 instead of or in addition to the computer-controlled machine 500 including the sensor 506. The control system 502 may also include the actuator 504 instead of or in addition to the computer-controlled machine 500 including the actuator 504.

[0071] like Figure 5As shown in , the control system 502 also includes a processor 520 and a memory 522. The processor 520 may include one or more processors. The memory 522 may include one or more memory devices. The classifier 514 (e.g., ML algorithm) of one or more embodiments may be implemented by the control system 502, which includes a non-volatile storage device 516, a processor 520, and a memory 522.

[0072] The non-volatile storage 516 may include one or more persistent data storage devices, such as a hard drive, an optical drive, a tape drive, a non-volatile solid-state device, a cloud storage device, or any other device capable of persistently storing information. The processor 520 may include one or more devices selected from a high-performance computing (HPC) system, including a high-performance core, a microprocessor, a microcontroller, a digital signal processor, a microcomputer, a central processing unit, a field programmable gate array, a programmable logic device, a state machine, a logic circuit, an analog circuit, a digital circuit, or any other device that manipulates (analog or digital) signals based on computer executable instructions residing in the memory 522. The memory 522 may include a single memory device or multiple memory devices, including but not limited to random access memory (RAM), volatile memory, non-volatile memory, static random access memory (SRAM), dynamic random access memory (DRAM), flash memory, cache memory, or any other device capable of storing information.

[0073] The processor 520 may be configured to read the memory 522 and execute computer executable instructions that reside in the non-volatile storage 516 and embody one or more ML algorithms and / or methods of one or more embodiments. The non-volatile storage 516 may include one or more operating systems and applications. The non-volatile storage 516 may store compilations and / or interpretations of computer programs created using a variety of programming languages ​​and / or technologies, including but not limited to Java, C, C++, C#, Objective C, Fortran, Pascal, JavaScript, Python, Perl, and PL / SQL, whether used alone or in combination.

[0074] When executed by the processor 520, the computer executable instructions of the non-volatile storage device 516 may cause the control system 502 to implement one or more ML algorithms and / or methods disclosed herein. The non-volatile storage device 516 may also include ML data (including data parameters) that support the functions, features, and processes of one or more embodiments described herein.

[0075] The program code embodying the algorithms and / or methods described herein can be distributed as a program product individually or collectively in a variety of different forms. The program code can be distributed using a computer-readable storage medium having a computer-readable program instruction thereon for enabling a processor to implement aspects of one or more embodiments. The computer-readable storage medium is essentially non-temporary and can include volatile and non-volatile, and removable and non-removable tangible media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. The computer-readable storage medium can also include RAM, ROM, erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other solid-state memory technology, portable compact disc read-only memory (CD-ROM) or other optical storage device, cassette, tape, disk storage device or other magnetic storage device, or any other medium that can be used to store desired information and can be read by a computer. The computer-readable program instruction can be downloaded from the computer-readable storage medium to a computer, another type of programmable data processing device, or another device, or downloaded to an external computer or external storage device via a network.

[0076] Computer-readable program instructions stored in a computer-readable medium can be used to instruct a computer, other types of programmable data processing devices, or other devices to operate in a particular manner, so that the instructions stored in the computer-readable medium produce an article of manufacture, including instructions for implementing the functions, actions, and / or operations specified in the flowchart or diagram. Consistent with one or more embodiments, in certain alternative embodiments, the functions, actions, and / or operations specified in the flowchart and diagram can be reordered, processed serially, and / or processed simultaneously. In addition, any flowchart and / or diagram may include more or fewer nodes or blocks than those illustrated in accordance with one or more embodiments.

[0077] The process, method or algorithm may be embodied in whole or in part using suitable hardware components, such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a state machine, a controller or other hardware components or devices, or a combination of hardware, software and firmware components.

[0078] Figure 6A schematic diagram of a control system 502 is depicted, which is configured to control a vehicle 600, which can be an at least partially autonomous vehicle or an at least partially autonomous robot. The vehicle 600 includes an actuator 504 and a sensor 506. The sensor 506 can include one or more video sensors, cameras, radar sensors, ultrasonic sensors, lidar sensors, and / or location sensors (e.g., GPS). One or more of the one or more specific sensors can be integrated into the vehicle 600. As an alternative or in addition to one or more specific sensors identified above, the sensor 506 can include a software module that is configured to determine the state of the actuator 504 when executed. A non-limiting example of a software module includes a weather information software module that is configured to determine the current or future state of the weather near the vehicle 600 or other location.

[0079] The classifier 514 of the control system 502 of the vehicle 600 may be configured to detect an object in the vicinity of the vehicle 600 depending on the input signal x. In such an embodiment, the output signal y may include information characterizing the vicinity of the object relative to the vehicle 600. The actuator control command 510 may be determined based on this information. The actuator control command 510 may be used to avoid a collision with the detected object.

[0080] In some embodiments, the vehicle 600 is an at least partially autonomous vehicle, and the actuator 504 can be embodied in the brakes, propulsion system, engine, transmission system, or steering device of the vehicle 600. Actuator control commands 510 can be determined so that the actuator 504 is controlled so that the vehicle 600 avoids colliding with the detected object. The detected object can also be classified according to what the classifier 514 thinks it is most likely to be (such as a pedestrian or a tree). The actuator control commands 510 can be determined depending on the classification. In the event of a possible adversarial attack, the system described above can also be trained to better detect objects or identify changes in lighting conditions or angles for sensors or cameras on the vehicle 600.

[0081] In some embodiments where the vehicle 600 is an at least partially autonomous robot, the vehicle 600 may be a mobile robot configured to perform one or more functions such as flying, swimming, diving, and stepping. The mobile robot may be an at least partially autonomous lawn mower or an at least partially autonomous cleaning robot. In such embodiments, actuator control commands 510 may be determined such that a propulsion unit, a steering unit, and / or a braking unit of the mobile robot may be controlled such that the mobile robot may avoid a collision with the identified object.

[0082] In some embodiments, vehicle 600 is an at least partially autonomous robot in the form of a gardening robot. In such embodiments, vehicle 600 may use an optical sensor as sensor 506 to determine the state of plants in the environment near vehicle 600. Actuator 504 may be a nozzle configured to spray a chemical. Depending on the identified plant species and / or the identified plant state, actuator control commands 510 may be determined to cause actuator 504 to spray a suitable amount of a suitable chemical onto the plant.

[0083] The vehicle 600 may be an at least partially autonomous robot in the form of a household appliance. Non-limiting examples of household appliances include a washing machine, a stove, an oven, a microwave, or a dishwasher. In such a vehicle 600, the sensor 506 may be an optical sensor configured to detect the state of an object to be processed by the household appliance. For example, in the case where the household appliance is a washing machine, the sensor 506 may detect the state of the clothes inside the washing machine. The actuator control command 510 may be determined based on the detected state of the clothes.

[0084] Figure 7 A schematic diagram of a control system 502 is depicted, the control system 502 being configured to control a system 700 (e.g., a manufacturing machine), such as a punch, cutter, or gun drill that manufactures the system 702 (such as part of a production line). The control system 502 may be configured to control an actuator 504 that is configured to control the system 700 (e.g., a manufacturing machine).

[0085] The sensor 506 of the system 700 (e.g., a manufacturing machine) may be an optical sensor configured to capture one or more properties of the manufactured product 704. The classifier 514 may be configured to determine a state of the manufactured product 704 based on the captured one or more properties. The actuator 504 may be configured to control the system 700 (e.g., a manufacturing machine) depending on the determined state of the manufactured product 704 so as to perform subsequent manufacturing steps of the manufactured product 704. The actuator 504 may be configured to control the function of the system 700 (e.g., a manufacturing machine) on the subsequent manufactured product 106 of the system 700 (e.g., a manufacturing machine) depending on the determined state of the manufactured product 704.

[0086] Figure 8 A schematic diagram of a control system 502 configured to control a power tool 800 , such as a drill or driver, having at least a partially autonomous mode is depicted. The control system 502 may be configured to control an actuator 504 configured to control the power tool 800 .

[0087] The sensor 506 of the power tool 800 can be an optical sensor configured to capture one or more properties of the working surface 802 and / or the fastener 804 driven into the working surface 802. The classifier 514 can be configured to determine the state of the working surface 802 and / or the fastener 804 relative to the working surface 802 based on the captured one or more properties. The state can be that the fastener 804 is flush with the working surface 802. Alternatively, the state can be the hardness of the working surface 802. The actuator 504 can be configured to control the power tool 800 so that the driving function of the power tool 800 is adjusted depending on the determined state of the fastener 804 relative to the working surface 802 or the one or more properties of the captured working surface 802. For example, if the state of the fastener 804 is flush with the working surface 802, the actuator 504 can stop the driving function. As another non-limiting example, the actuator 504 can apply additional or less torque depending on the hardness of the working surface 802.

[0088] Fig. 9 A schematic diagram of a control system 502 is depicted, which is configured to control an automated personal assistant 900. The control system 502 may be configured to control an actuator 504, which may be configured to control the automated personal assistant 900. The automated personal assistant 900 may be configured to control a household appliance, such as a washing machine, a stove, an oven, a microwave, or a dishwasher.

[0089] Sensor 506 may be an optical sensor and / or an audio sensor. The optical sensor may be configured to receive a video image of user 902's gesture 904. The audio sensor may be configured to receive a voice command of user 902.

[0090] The control system 502 of the automated personal assistant 900 may be configured to determine an actuator control command 510 that is configured to control the system 502. The control system 502 may be configured to determine the actuator control command 510 based on the sensor signal 508 of the sensor 506. The automated personal assistant 900 is configured to transmit the sensor signal 508 to the control system 502. The classifier 514 of the control system 502 may be configured to execute a gesture recognition algorithm to identify a gesture 904 made by the user 902 in order to determine the actuator control command 510 and transmit the actuator control command 510 to the actuator 504. The classifier 514 may be configured to retrieve information from the non-volatile storage device in response to the gesture 904 and output the retrieved information in a form suitable for receipt by the user 902.

[0091] Fig.10A schematic diagram of a control system 502 is depicted that is configured to control a monitoring system 1000. The monitoring system 1000 can be configured to physically control access through a door 1002. A sensor 506 can be configured to detect scenes relevant to a decision whether to grant access. The sensor 506 can be an optical sensor that is configured to generate and transmit image and / or video data. The control system 502 can use such data to detect a person's face.

[0092] The classifier 514 of the control system 502 of the monitoring system 1000 can be configured to interpret the image and / or video data by matching the identities of known persons stored in the non-volatile storage device 516 to determine the identity of the person. The classifier 514 can be configured to generate an actuator control command 510 in response to the interpretation of the image and / or video data. The control system 502 can be configured to transmit the actuator control command 510 to the actuator 504. In this embodiment, the actuator 504 can be configured to lock or unlock the door 1002 in response to the actuator control command 510. In some embodiments, non-physical logical access control is also possible.

[0093] Monitoring system 1000 may also be a surveillance system. In such an embodiment, sensor 506 may be an optical sensor configured to detect a monitored scene, and control system 502 is configured to control display 1004. Classifier 514 is configured to determine a classification of a scene, for example, whether a scene detected by sensor 506 is suspicious. Control system 502 is configured to transmit actuator control commands 510 to display 1004 in response to the classification. Display 1004 may be configured to adjust displayed content in response to actuator control commands 510. For example, display 1004 may highlight an object deemed suspicious by classifier 514. Using embodiments of the disclosed system, the surveillance system may predict the appearance of an object at certain times in the future.

[0094] Fig.11 A schematic diagram of a control system 502 is depicted, the control system 502 being configured to control an imaging system 1100, such as an MRI device, an x-ray imaging device, or an ultrasound device. The sensor 506 may be, for example, an imaging sensor. The classifier 514 may be configured to determine a classification of all or part of a sensed image. The classifier 514 may be configured to determine or select an actuator control command 510 in response to a classification obtained by a trained neural network. For example, the classifier 514 may interpret an area of ​​the sensed image as a potential anomaly. In this case, the actuator control command 510 may be determined or selected to cause the display 1102 to display the imaging and highlight the area of ​​the potential anomaly.

[0095] In some embodiments, a method for data enhancement includes receiving audio stream data associated with at least one pulse event, receiving a label associated with the audio stream data, and detecting at least one peak of the at least one pulse event using a starting point detector. The method also includes extracting at least one positive sample of the audio stream data associated with the at least one pulse event, wherein the at least one positive sample includes a portion of the audio stream data associated with a first time value of the audio stream data. The method also includes applying the label associated with the audio stream data to the at least one positive sample, and extracting at least one negative sample of the audio stream data associated with the at least one pulse event, wherein the at least one negative sample does not include a portion of the audio stream data associated with the first time value of the audio stream data. The method also includes enhancing training data based on the at least one positive sample and the at least one negative sample, and training at least one machine learning model using the enhanced training data.

[0096] In some embodiments, at least one positive sample extends a first time period before the first time value and extends a second time period after the first time value. In some embodiments, at least one negative sample includes a label different from a label associated with the audio stream data. In some embodiments, at least one negative sample and at least one positive sample do not overlap. In some embodiments, at least one machine learning model is configured to detect at least a class of sounds associated with at least one impulse event. In some embodiments, training at least one machine learning model using enhanced training data includes training at least one machine learning model using at least one back propagation technique. In some embodiments, the method also includes enhancing the training data using at least one of the following: spectral masking, spectral shift, at least one background sound, at least one impulse response associated with an acoustic environment. In some embodiments, the starting point detector is configured to detect at least one peak using at least one of the following: an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data.

[0097] In some embodiments, a system for data enhancement includes a processor and a memory. The memory includes instructions that, when executed by the processor, cause the processor to: receive audio stream data associated with at least one pulse event; receive a label associated with the audio stream data; detect at least one peak of at least one pulse event using a starting point detector; extract at least one positive sample of the audio stream data associated with the at least one pulse event, wherein the at least one positive sample includes a portion of the audio stream data associated with a first time value of the audio stream data; apply the label associated with the audio stream data to the at least one positive sample; extract at least one negative sample of the audio stream data associated with the at least one pulse event, wherein the at least one negative sample does not include a portion of the audio stream data associated with the first time value of the audio stream data; enhance training data based on the at least one positive sample and the at least one negative sample; train at least one machine learning model using the enhanced training data; and use the at least one machine learning model to detect one or more pulse events associated with the audio data provided as input to the at least one machine learning model.

[0098] In some embodiments, at least one positive sample extends a first time period before the first time value and extends a second time period after the first time value. In some embodiments, at least one negative sample includes a label different from the label associated with the audio stream data. In some embodiments, at least one negative sample and at least one positive sample do not overlap. In some embodiments, at least one machine learning model is configured to detect at least a class of sounds associated with at least one impulse event. In some embodiments, training at least one machine learning model using enhanced training data includes training at least one machine learning model using at least one back propagation technique. In some embodiments, the instructions also cause the processor to further enhance the training data using at least one of the following: spectral masking, spectral shift, at least one background sound, at least one impulse response associated with an acoustic environment. In some embodiments, the starting point detector is configured to detect at least one peak using at least one of the following: an energy envelope associated with the audio stream data, at least one energy lower limit associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data.

[0099] In some embodiments, an apparatus for data enhancement includes a processor and a memory. The memory includes instructions that, when executed by the processor, cause the processor to: receive audio stream data associated with at least one pulse event; receive a label associated with the audio stream data; use a start detector to detect at least one peak of the at least one pulse event, wherein the start detector is configured to detect the at least one peak using at least one of the following: an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data; extract at least one of the audio stream data associated with the at least one pulse event; positive samples, wherein the at least one positive sample includes a portion of the audio stream data associated with a first time value of the audio stream data; applying a label associated with the audio stream data to the at least one positive sample; extracting at least one negative sample of the audio stream data associated with the at least one impulse event, wherein the at least one negative sample does not include the portion of the audio stream data associated with the first time value of the audio stream data; based on the at least one positive sample and the at least one negative sample, enhancing the training data using at least one of the following: spectral masking, spectral shifting, at least one background sound, at least one impulse response associated with an acoustic environment; and training at least one machine learning model using the enhanced training data.

[0100] In some embodiments, at least one positive sample extends a first period of time before the first time value and extends a second period of time after the first time value. In some embodiments, at least one negative sample includes a label that is different from a label associated with the audio stream data. In some embodiments, at least one negative sample and at least one positive sample do not overlap.

[0101] In some embodiments, a computer-implemented method for data augmentation includes receiving audio stream data associated with at least one impulse event, receiving a label associated with the audio stream data, and detecting at least one peak of the at least one impulse event using an onset detector algorithm. The method also includes extracting at least one positive sample of the audio stream data associated with the at least one impulse event, wherein the at least one positive sample includes a portion of the audio stream data associated with a first time value of the audio stream data. The method also includes labeling the at least one positive sample with the label associated with the audio stream data to produce at least one labeled positive sample, and extracting at least one negative sample of the audio stream data associated with the at least one impulse event, wherein the at least one negative sample does not include the portion of the audio stream data associated with the first time value of the audio stream data. The method also includes augmenting machine learning training data based on the at least one labeled positive sample and the at least one negative sample.

[0102] In some embodiments, at least one positive sample extends a first time period before the first time value and extends a second time period after the first time value. In some embodiments, at least one negative sample includes a label different from the label associated with the audio stream data. In some embodiments, at least one negative sample and at least one positive sample do not overlap. In some embodiments, the method also includes training at least one machine learning model with enhanced machine learning training data, wherein the at least one machine learning model is configured to at least detect a class of sounds associated with the at least one impulse event. In some embodiments, training at least one machine learning model with enhanced machine learning training data includes training at least one machine learning model using at least one back propagation technique. In some embodiments, the method also includes further enhancing the machine learning training data using at least one of the following: spectral masking, spectral shift, at least one background sound, at least one impulse response associated with an acoustic environment. In some embodiments, the starting point detector algorithm is configured to detect at least one peak using at least one of the following: an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data.

[0103] In some embodiments, a system for training a pulse event detection machine learning model includes a processor and a memory. The memory includes instructions that, when executed by the processor, cause the processor to: receive audio stream data associated with at least one pulse event; receive a label associated with the audio stream data; detect at least one peak of at least one pulse event using an onset detector algorithm; extract at least one positive sample of the audio stream data associated with the at least one pulse event, wherein the at least one positive sample includes a portion of the audio stream data associated with a first time value of the audio stream data; label the at least one positive sample with the label associated with the audio stream data to produce at least one labeled positive sample; extract at least one negative sample of the audio stream data associated with the at least one pulse event, wherein the at least one negative sample does not include a portion of the audio stream data associated with the first time value of the audio stream data; enhance the machine learning training data based on the at least one labeled positive sample and the at least one negative sample; and train the pulse event detection machine learning model using the enhanced machine learning training data.

[0104] In some embodiments, at least one positive sample extends a first time period before the first time value and extends a second time period after the first time value. In some embodiments, at least one negative sample includes a label different from the label associated with the audio stream data. In some embodiments, at least one negative sample and at least one positive sample do not overlap. In some embodiments, the instructions also cause the processor to use an impulse event detection machine learning model to detect one or more impulse events associated with the audio data provided as input to at least one machine learning model for impulse event detection. In some embodiments, training the impulse event detection machine learning model using enhanced machine learning training data includes training the impulse event detection machine learning model using at least one back propagation technique. In some embodiments, the instructions also cause the processor to further enhance the machine learning training data using at least one of the following: spectral masking, spectral shift, at least one background sound, at least one impulse response associated with an acoustic environment. In some embodiments, the starting point detector algorithm is configured to detect at least one peak using at least one of the following: an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data.

[0105] In some embodiments, an apparatus for pulse event detection includes a processor and a memory. The memory includes instructions that, when executed by the processor, cause the processor to: receive audio stream data associated with at least one pulse event; receive a label associated with the audio stream data; use a starting point detector algorithm to detect at least one peak of the at least one pulse event, wherein the starting point detector is configured to detect the at least one peak using at least one of the following: an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data; extract at least one positive sample of the audio stream data associated with the at least one pulse event, wherein the at least one positive sample includes audio stream data associated with a first time value of the audio stream data; a portion of the audio stream data; marking at least one positive sample with a label associated with the audio stream data to produce at least one labeled positive sample; extracting at least one negative sample of the audio stream data associated with the at least one pulse event, wherein the at least one negative sample does not include a portion of the audio stream data associated with a first time value of the audio stream data; based on the at least one labeled positive sample and the at least one negative sample, enhancing the machine learning training data using at least one of the following: spectral masking, spectral shift, at least one background sound, at least one impulse response associated with an acoustic environment; training at least one machine learning model using the enhanced machine learning training data; and using the at least one machine learning model to detect one or more impulse events associated with the audio data provided as input to the at least one machine learning model.

[0106] In some embodiments, at least one positive sample extends a first period of time before the first time value and extends a second period of time after the first time value. In some embodiments, at least one negative sample includes a label that is different from a label associated with the audio stream data. In some embodiments, at least one negative sample and at least one positive sample do not overlap.

[0107] The process, method or algorithm disclosed herein can be transferred to / implemented by a processing device, a controller or a computer, which can include any existing programmable electronic control unit or a dedicated electronic control unit. Similarly, the process, method or algorithm can be stored in many forms as data and instructions that can be executed by a controller or a computer, including but not limited to information permanently stored on a non-writable storage medium such as a ROM device and information that can be alternatively stored on a writable storage medium such as a floppy disk, a tape, a CD, a RAM device and other magnetic and optical media. The process, method or algorithm can also be embodied in a software executable object. Alternatively, suitable hardware components, such as an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), a state machine, a controller or other hardware components or devices, or a combination of hardware, software and firmware components, can be used to embody these processes, methods or algorithms in whole or in part.

[0108] Although exemplary embodiments are described above, these embodiments are not intended to describe all possible forms covered by the claims. The words used in the specification are descriptive rather than restrictive, and it will be understood that various changes can be made without departing from the spirit and scope of the present disclosure. As described above, the features of the various embodiments can be combined to form other embodiments of the present invention that may not be explicitly described or illustrated. Although various embodiments may have been described as providing advantages or being superior to other embodiments or prior art implementations with respect to one or more desired characteristics, it will be recognized by those of ordinary skill in the art that one or more features or characteristics can be compromised to achieve the desired overall system properties, depending on the specific application and implementation. These properties may include, but are not limited to, cost, strength, durability, life cycle cost, marketability, appearance, packaging, size, applicability, weight, manufacturability, ease of assembly, etc. Thus, to the extent that any embodiment is described as less desirable than other embodiments or prior art implementations with respect to one or more characteristics, then these embodiments do not exceed the scope of the present disclosure and may be desirable for specific applications.

Claims

1. A computer-implemented method for data augmentation, the method comprising: receiving audio stream data associated with at least one pulse event; Receiving a tag associated with the audio stream data; detecting at least one peak of the at least one pulse event using an onset detector algorithm; extracting at least one positive sample of the audio stream data associated with the at least one impulse event, wherein the at least one positive sample comprises a portion of the audio stream data associated with a first time value of the audio stream data; labeling the at least one positive sample using the label associated with the audio stream data to generate at least one labeled positive sample; extracting at least one negative sample of the audio stream data associated with the at least one impulse event, wherein the at least one negative sample does not include a portion of the audio stream data associated with the first time value of the audio stream data; as well as The machine learning training data is enhanced based on the at least one labeled positive sample and the at least one negative sample. 2 . The computer-implemented method of claim 1 , wherein the at least one positive sample extends a first period of time before the first time value and extends a second period of time after the first time value. 3 . The computer-implemented method of claim 1 , wherein the at least one negative sample comprises a label different from the label associated with the audio stream data. The computer-implemented method of claim 1 , wherein the at least one negative sample and the at least one positive sample do not overlap.

5. The computer-implemented method according to claim 1 further includes training at least one machine learning model with enhanced machine learning training data, wherein the at least one machine learning model is configured to at least detect a type of sound associated with the at least one pulse event.

6. A computer-implemented method according to claim 5, wherein training the at least one machine learning model with the enhanced training data includes training the at least one machine learning model using at least one back-propagation technique.

7. The computer-implemented method of claim 1, further comprising further enhancing the machine learning training data using at least one of the following: spectral masking, spectral shifting, at least one background sound, at least one impulse response associated with an acoustic environment.

8. The computer-implemented method of claim 1 , wherein the onset detector algorithm is configured to detect the at least one peak using at least one of: an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data.

9. A system for training a machine learning model for pulse event detection, the system comprising: processor; as well as a memory comprising instructions which, when executed by the processor, cause the processor to: receiving audio stream data associated with at least one pulse event; Receiving a tag associated with the audio stream data; detecting at least one peak of the at least one pulse event using an onset detector algorithm; extracting at least one positive sample of the audio stream data associated with the at least one impulse event, wherein the at least one positive sample comprises a portion of the audio stream data associated with a first time value of the audio stream data; labeling the at least one positive sample using the label associated with the audio stream data to generate at least one labeled positive sample; extracting at least one negative sample of the audio stream data associated with the at least one impulse event, wherein the at least one negative sample does not include a portion of the audio stream data associated with the first time value of the audio stream data; enhancing machine learning training data based on the at least one labeled positive sample and the at least one negative sample; as well as The pulse event detection machine learning model is trained using the enhanced machine learning training data. 10 . The system of claim 9 , wherein the at least one positive sample extends a first period of time before the first time value and extends a second period of time after the first time value.

11. The system of claim 9, wherein the at least one negative sample comprises a label different from the label associated with the audio stream data.

12. The system of claim 9, wherein the at least one negative sample and the at least one positive sample do not overlap.

13. A system according to claim 9, wherein the instructions further cause the processor to use the pulse event detection machine learning model to detect one or more pulse events associated with audio data provided as input to the pulse event detection machine learning model.

14. The system of claim 9, wherein using the enhanced machine learning training data to train the pulse event detection machine learning model comprises utilizing at least one back-propagation technique to train the pulse event detection machine learning model.

15. The system of claim 9, wherein the instructions further cause the processor to further enhance the machine learning training data using at least one of: spectral masking, spectral shifting, at least one background sound, at least one impulse response associated with an acoustic environment.

16. The system of claim 9, wherein the onset detector algorithm is configured to detect the at least one peak using at least one of: an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data.

17. An apparatus for pulse event detection, the apparatus comprising: processor; as well as a memory comprising instructions which, when executed by a processor, cause the processor to: receiving audio stream data associated with at least one pulse event; Receiving a tag associated with the audio stream data; detecting at least one peak of the at least one impulse event using an onset detector algorithm, wherein the onset detector algorithm is configured to detect the at least one peak using at least one of: an energy envelope associated with the audio stream data, at least one energy floor associated with the audio stream data, at least one energy peak associated with the audio stream data, and at least one dynamic range associated with the audio stream data; extracting at least one positive sample of the audio stream data associated with the at least one impulse event, wherein the at least one positive sample comprises a portion of the audio stream data associated with a first time value of the audio stream data; labeling the at least one positive sample using the label associated with the audio stream data to generate at least one labeled positive sample; extracting at least one negative sample of the audio stream data associated with the at least one impulse event, wherein the at least one negative sample does not include a portion of the audio stream data associated with the first time value of the audio stream data; Based on the at least one labeled positive sample and the at least one negative sample, enhancing the machine learning training data using at least one of: spectral masking, spectral shifting, at least one background sound, at least one impulse response associated with an acoustic environment; training at least one machine learning model using the augmented machine learning training data; as well as The at least one machine learning model is used to detect one or more pulse events associated with audio data provided as input to the at least one machine learning model.

18. The apparatus of claim 17, wherein the at least one positive sample extends a first period of time before the first time value and extends a second period of time after the first time value.

19. The apparatus of claim 17, wherein the at least one negative sample comprises a label different from the label associated with the audio stream data.

20. The apparatus of claim 17, wherein the at least one negative sample and the at least one positive sample do not overlap.

Images