Aspects of the present disclosure relate to stress compensation or image sticking or ghosting effect in a display device.
Compensation for output decline in a video display such as an organic light-emitting diode (OLED) display may be used to preserve image quality as a display ages. The data used to perform such compensation may be stored in compressed form to reduce memory requirements; however, errors in such compressed data may accumulate unevenly, resulting in loss of image quality.
Additionally, when an OLED display device displays a static image for a prolonged period of time, it may suffer from image retention. The result is that once the static image is removed or changed, a faint outline—or ghosting—of the original image may still be visible to a user, even when the image content has changed. This is often referred to as image sticking, image retention, or image ghosting. The same method as stress profile is applied for image sticking.
Thus, there is a need for an improved system and method for stress compensation to reduce or eliminate output decline of OLED display devices.
The above information disclosed in this Background section is only for enhancement of understanding of the background of the disclosure and therefore it may contain information that does not constitute prior art.
This summary is provided to introduce a selection of features and concepts of embodiments of the present disclosure that are further described below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used in limiting the scope of the claimed subject matter. One or more of the described features may be combined with one or more other described features to provide a workable device.
Aspects of example embodiments of the present disclosure relate to a system and a method for reducing or eliminating display output decline by mitigating the effects of pixel degradation/aging in a display device.
Aspects of example embodiments of the present disclosure relate to a system and method utilizing dither to compensate for quantization noise in iterative compression on additive data.
According to some embodiments of the present disclosure, there is provided a method of stress compensation in a display device, the method including: retrieving, by a decoder, compressed stress data that is quantized by a quantization value; decoding, by the decoder, the compressed stress data to generate decoded data; generating a dither value based on the quantization value; and adding the dither value to the decoded data to compensate for quantization of the compressed stress data.
In some embodiments, the generating and the adding are performed by the decoder or a circuit external to the decoder.
In some embodiments, the quantization value is a power of two, the quantization value is stored in a header of the compressed stress data, and the compressed stress data corresponds to brightness values of a slice of a frame of pixel data to be displayed on a display device.
In some embodiments, the generating the dither value is further based on a previous dither value corresponding to the compressed stress data, wherein the previous dither value is stored in a header of the compressed stress data.
In some embodiments, the generating the dither value includes utilizing a uniform pseudo-random number generator.
In some embodiments, the generating the dither value includes: determining, by the decoder, that the quantization value is equal to 1; and in response, outputting, by the decoder, the dither value as zero.
In some embodiments, the generating the dither value includes: determining, by the decoder, that the quantization value is equal to 2; and in response, outputting, by the decoder, the dither value as equaling a modulus of a previous dither value plus one and the quantization value.
In some embodiments, the generating the dither value includes: outputting, by the decoder, the dither value as equaling a modulus of a first value and the quantization value, wherein the first value is calculated as an addition of a previous dither value and half of the quantization value subtracted by one.
In some embodiments, the method of further includes: storing, by the decoder, the dither value in a header of the decoded data.
According to some embodiments of the present disclosure, there is provided a method of stress compensation in a display device, the method including: compressing, by an encoder, a first accumulated stress data at a quantization corresponding to a quantization value to generate compressed accumulated stress data; storing, by the encoder, the compressed accumulated stress data in a memory; retrieving and decompressing, by a decoder, the compressed accumulated stress data to generate decoded data; generating, by the decoder, a dither value based on the quantization value; and adding, by the decoder, the dither value to the decoded data to compensate for the quantization of the first accumulated stress data.
In some embodiments, the method further includes: receiving, by a stress capture module, an output image to be displayed on the display device; and calculating, by the stress capture module, second stress data based on the output image.
In some embodiments, the method further includes: receiving, by an adding circuit, second stress data corresponding to a frame of image data to be displayed on the display device; and adding, by the adding circuit, the decoded data to the second stress data to generate an updated accumulated stress data.
In some embodiments, the method further includes: compressing, by the encoder, the updated accumulated stress data for storage in the memory.
In some embodiments, the generating the dither value is further based on a previous dither value corresponding to the first accumulated stress data, and wherein the previous dither value corresponds to a slice of a frame of pixel data different from a slice of the frame of pixel data corresponding to the second stress data.
In some embodiments, the first accumulated stress data corresponds to brightness values of a slice of a frame of pixel data to be displayed on a display device.
In some embodiments, the generating the dither value is further based on a previous dither value corresponding to the first accumulated stress data, and wherein the quantization value and the previous dither value are stored in a header of the compressed accumulated stress data.
In some embodiments, the generating the dither value includes: determining, by the decoder, the quantization value; in response to determining that the quantization value is equal to 1, outputting, by the decoder, the dither value as zero; and in response to determining that the quantization value is equal to 2, outputting, by the decoder, the dither value as equaling a modulus of a previous dither value plus one and the quantization value.
In some embodiments, the generating the dither value includes: outputting, by the decoder, the dither value as equaling a modulus of a first value and the quantization value, and wherein the first value is calculated as an addition of a previous dither value and half of the quantization value subtracted by one.
In some embodiments, the method further includes: storing, by the decoder, the dither value in a header of the decoded data.
According to some embodiments of the present disclosure, there is provided a system for performing stress compensation in a display device, the system including: a memory; and a processing circuit configured to perform: retrieving compressed stress data that is quantized by a quantization value; decoding the compressed stress data to generate decoded data; generating a dither value based on the quantization value; and adding the dither value to the decoded data to compensate for quantization of the compressed stress data.
These and other features of some example embodiments of the present disclosure will be appreciated and understood with reference to the specification, claims, and appended drawings, wherein:
The detailed description set forth below in connection with the appended drawings is intended as a description of some example embodiments of a system and a method for mitigating the effects of compression errors provided in accordance with the present disclosure and is not intended to represent the only forms in which the present disclosure may be constructed or utilized. The description sets forth the features of the present disclosure in connection with the illustrated embodiments. It is to be understood, however, that the same or equivalent functions and structures may be accomplished by different embodiments that are also intended to be encompassed within the scope of the disclosure. As denoted elsewhere herein, like element numbers are intended to indicate like elements or features.
Certain kinds of display devices may have characteristics that change with use. For example, an organic light-emitting diode (OLED) display device may include a display panel having a plurality of pixels, each including several subpixels (e.g., a red subpixel, a green subpixel, and a blue subpixel), and each of the subpixels may include an organic light-emitting diode configured to emit a different respective color. Each organic light-emitting diode may have an optical efficiency that declines with use, so that, for example, after the organic light-emitting diode has been in operation for some time, the optical output at a certain current may be lower than it was, at the same current, when the organic light-emitting diode was new.
This reduction in optical efficiency may result in dimming of parts of a display panel that have on average, during the life of the display device, displayed brighter portions of the displayed images than other parts of the display. For example, a display device used to view largely unchanging images from a security camera, the field of view of which contains a scene having a first portion that is sunlit, and relatively bright, during most of the day, and a second portion that is in the shade and relatively dim, during most of the day, may eventually show a more significant decrease in optical efficiency in the first portion than in the second portion. The fidelity of image reproduction of such a display device may degrade over time as a result. As another example, a display device that is used part of the time to display white text at the bottom of the image, separated by a black margin from the rest of the image, may experience a lower reduction of optical efficiency in the black margin than in other parts of the display panel, so that if the display device is later used in a mode in which a scene fills the entire display panel, a brighter band may appear where the black margin was previously displayed. This may be referred to as image sticking or ghosting.
To reduce the effect of such non-uniformities in the optical efficiency of a display device, a display device may include features to compensate for the reduction of optical efficiency resulting from use of the display.
Referring to
In some example embodiments, the method of determining the “stress profile” may be a data driven method to compensate for OLED output decline. In some example embodiments, in order to determine the “stress profile” of the display device 100, the memory 120 may accumulate one number for one or some sub-pixels, each time a new image is displayed in the display device 100 as a new entry in the “stress profile” or “stress table” for the display device 100. For example, as a continuous stream of images together form displayed video in the display device 100, the drive current for each sub-pixel in each image may be measured and a number indicating the current or brightness of the subpixel may be added to the respective number or “stress” for that sub-pixel in “stress profile” or stress table” in the memory 120. In some examples, the stress value for each sub-pixel may be calculated as the total sum of the brightness values displayed by the sub-pixel.
In some example embodiments, a display device 100 includes a timing controller and a plurality of driver integrated circuits (e.g., a scan/gate driver and a data driver). The processing circuit 115 may be, or may be part of, one or more of the driver integrated circuits. In some embodiments, each driver integrated circuit is responsible for driving a portion of the display panel 110, and it may accordingly perform stress accumulation and stress compensation for that portion, independently of the other driver integrated circuits.
In some example embodiments, during operation of the display device 100, the drive current to each sub-pixel may be adjusted to compensate for an estimated loss of optical efficiency. The estimated loss of optical efficiency may be based on the lifetime stress of the sub-pixel. For example, the drive current to each sub-pixel may be increased in accordance with (e.g., in proportion to) the estimated loss of optical efficiency of the sub-pixel (e.g., as accumulated in the memory 120), so that the optical output of the sub-pixel may be substantially the same as it would have been had the optical efficiency of the sub-pixel not been reduced, and had the drive current not been increased. In some example embodiments, a non-linear function based on empirical data or a model of the physics of the sub-pixel may be used to infer or predict the loss of optical efficiency expected to be present, based on the lifetime stress of the sub-pixel. The calculations of the predicted loss of optical efficiency, and of the accordingly adjusted drive current, may be performed by the processing circuit 115. In some embodiments, the calculations are stored in a look up table in the memory and the processing circuit 115 may use it for compensation. In some example embodiments, during operation of the display device 100, the drive current to each sub-pixel may be adjusted according to a compensation factor, which may be calculated based on an accumulated brightness captured from output image of the sub-pixel.
In some example embodiments, tracking the total stress of each sub-pixel may require a significant amount of memory. For example, for a display with 1920×1080 pixels, with three sub-pixels per pixel, and with the stress of each sub-pixel stored in a few byte numbers, the size of the memory required may be in the order of tens of megabytes. Moreover, the computational burden of updating each stress number for each sub-pixel for each frame of video (e.g., for each displayed image) is significant.
Various approaches may also be used to reduce the memory size required for storing sub-pixel stress in the stress table. In some embodiments, the memory on the stress profile chipset is reduced by compressing the data stored in the memory.
According to some embodiments, the system 200 also includes a first decoder 230a, an encoder 235, and a second decoder 230b. In some example embodiments, a compressed representation of the stress table is stored in the memory 205. In operation, compressed stress data may be accessed (e.g., read out) via the memory controller 225 and may be decompressed by the first decoder 230a before being fed to the drive current adjustment circuit of the compensation module 210. The drive current adjustment circuit of the compensation module 210 calculates the adjusted drive current values for each sub-pixel based on the respective compensation factor of that sub-pixel. The compensation factor for each sub-pixel is based on the stress value of that sub-pixel. The adjusted drive current value for each sub-pixel is read by the stress capture module 215. The adjusted drive current value for each sub-pixel represents the current rate of accumulation of stress of the sub-pixels being displayed. The compressed stress data in the memory 205 are also decompressed by a second decoder 230b to retrieve each previously stored (e.g., in the memory 205) stress value for each sub-pixel. The decompressed stress values for the sub-pixel from the second decoder 230b are sent to the adding circuit 220. Each previously stored stress value for each sub-pixel is increased (or “augmented”) in the adding circuit 220 based on the current rate of accumulation of stress (e.g., by a number proportional to the adjusted drive current value). The augmented stress values from the adding circuit 220 are compressed by the encoder 235, before being stored in the memory 205. The encoder 235 compresses data that it receives in a manner that reduces the size of the stored data. In some examples, the compression applied by the compressor may be lossy, in order to reduce (e.g., minimize) the amount of memory capacity consumed by the compressed data. Each of the first decoder 230a and the second decoder 230b decompresses the received data. For example, each of the first decoder 230a and the second decoder 230b performs an operation that inverts, or approximately inverts, the operation performed by the encoder 235. Various methods of compression may be employed, including entropy coding, such as Huffman coding or arithmetic coding.
In some examples, the output of first decoder 230a may be truncated so as to discard the lesser significant bits of the decoded value. This may be done due to the fact that while the lesser significant bits serve to accurately measure the accumulated stress, they may not materially affect the compensation performed by the compensation module 210.
According to some embodiments, the encoder and decoders operate on a unit of data referred to as a slice. The slice is an independent coding unit of image data that represents a portion of the image. For example, the slice may be a portion that spans the entire width and four vertical lines of the image. Each iteration of the encoding/decoding process described above may be performed on one slice of the image data.
While the first and second decoders 230a and 230b have been illustrated in
The system 200 of
Referring to
In some embodiments, the decoder 230 includes an entropy decoder 315, a dither generator 320, and an adder 325. The entropy decoder 315 may be a lossless decoder (e.g., an arithmetic or Huffman decoder) that performs the inverse operation of the entropy encoder 310. In some embodiments, the entropy decoder 315 includes a scaler configured to multiply its input by 2m, which is the quantization value utilized by the encoder 235. The adder 335 adds the outputs of the entropy decoder 315 and the dither generator 320.
In effect, the combined operation of the quantizer 305 and the scaler may be formulaically expressed by Equation (1):
where x represents an input value, such as the retrieved stress value, m represents the exponential quantization parameter,
represents the bit operation performed by the quantizer 305 of the encoder 235, the multiplication by 2m is performed by the scaler of the decoder 230, and q′m(x) represents the quantized value of x.
The quantization and scaling operations have the property that can be expressed by Equation 2:
q′m(q′m(x))=q′m(x) Equation (2)
In other words, consecutive iterations of the application of the quantization and scaling does not add further error. That is, the codec device is idempotent (i.e., successive operation of the decoder-encoder operations does not change the results beyond the initial application) or substantially idempotent.
According to some embodiments, the dither generator 320 generates and outputs a dither (e.g., a quantized-constrained dither) that is added to the output of the entropy decoder 315 via the adder 325. The dither may be a uniform dither calculated based on the quantization value n=2m. The dither values may be {0, 1 . . . n−1}. As will be described below, due to the nature of the encoding algorithm, the added dither results in unbiased addition of stress values (i.e., an addition that does not accumulate quantization error).
In some embodiments, the same dither value is added to the decompressed stress values of all of the pixels in the same slice. This ensures that additional entropy is not introduced through the next addition operation of the next iteration of the codec device. This is because the added dither is more useful temporally (per iteration) rather than spatially. If the dither were different spatially for different pixels, it would unnecessarily introduce additional noise which is more difficult to compress. For example, if different dither values are used within the same slice, brightness values of pixels in a flat area may not be the same after further addition and quantizing. In some examples, the dither varies slice to slice, even for the same quantization value, as each slice is treated independently and the start of the dither sequence is based on when the quantization changes and is independent from the other slices. In effect, the dither allows the codec to effectively subsample the addition performed by the adding circuit 220 while maintaining unbiased estimation. This can be mathematically illustrated as follows.
It can also be shown that where
E[qn(X+c)]=c Equation (3)
where n is the quantization value 2m, c is a non-negative integer representing the stress value of a single subpixel (e.g., a red, green, or blue subpixel of a pixel) for a given iteration calculated by the stress capture module 215, X is a random variable between 0 and (n−1) representing the dither value, and E[qn(X+c)] is the expected value of the quantized stress value, which may be stored at the memory 205. As an example, if the quantization value n is equal to 4, and c is equal to 1, without the dither (i.e., without the random variable X), the expected value described by Equation (3) is 0; however, with dither, the expected value is c=1. In other words, with the dither added, the expected value of the quantized/compressed stress value being stored in memory is the same as the uncompressed stress value.
Further, it can be shown that:
E[qn
where Xn
Equation (4) is an expression of two the expected value of the stored stress value for two consecutive iterations. However, even when expanded to an arbitrary number of iterations, the overall conclusion still holds that the expected value of the compressed accumulated stress value, according to some embodiments, is the same as the accumulated value in uncompressed form. This is due to the application of the dither that is set based on the quantization value 2m. Thus, the added dither may compensate for the quantization effect of the encoder 235. As a result, the encoder 235 and decoder 230 combination may function as an unbiased estimator for compound addition of stress values.
In some examples, the encoder 235 may store the quantization value in the header of the encoded data and the decoder 230 may recover the dither value in the header of the decoded data. Thus, it may be unnecessary to keep track of the iteration number. In this manner, the decoder 230 can always identify the previous quantization and dither values, which the decoder 230 may use to determine the dither value in the current iteration.
According to some embodiments, when the decoder 230 determines that no quantization was applied in the previous iteration (e.g., determines that the exponential quantization parameter m=0 or that the quantization value 2m=1), the dither value is set to zero and no dither is applied to the encoded data. When the decoder 230 determines the quantization value equals 2 (i.e., m=1), the decoder 230 calculates the current dither value as modulus, or remainder after integer division of, the previous dither value plus one and the quantization value. Otherwise (i.e., m>1), the decoder 230 calculates the dither value as the modulus of the addition of the previous dither value and 2m-1−1 and the quantization value. While a slice of data may use a single dither value, it can be proven that by using a modulo additive value of 2m-1−1, all dither values will be used (over many iterations) for a given quantization value, as long as the quantization step doesn't change between iterations. If not all the values are used, it is possible that the result from the addition and quantization steps may be biased from the actual value over many iterations. In general, the dither sequence needs to only conform to a uniform pseudo random number sequence based on the quantization step size to support unbiased iterative addition. The above description is just one possible approximation which does not require extra information in the header to hold the state of the psuedo-random number (PRN) generator; and embodiments of the present invention are not limited thereto. For example, rather than “bouncing” around all of dither values, as done above, one may progress through the possible dither values in a linear fashion.
The calculation of the dither value based on the quantization value and the previous dither value (i.e., the dither value of the previous iteration), according to some embodiments, may also be expressed in pseudo code as:
Thus, according to some embodiments, by using the quantization-constrained dither, the codec can compress/quantize the accumulated stress value data without additive bias. This also allows for the changes in quantization to dynamically match the memory requirements. The method and system, according to some embodiments of the present disclosure, may be readily implemented as it does not require knowledge of the output values of the decoder or the previous values input to the encoder.
Referring to
Referring to
The stress capture module 215 may receive an output image to be displayed on the display device 100, and calculate second stress data based on the output image. The adding circuit 220 may receive the second stress data corresponding to a frame of image data to be displayed on the display device 100, and add the decoded data to the second stress data to generate an updated accumulated stress data. The encoder 235 may compress the updated accumulated stress data for storage in the memory 205. The previous dither value may be used in conjunction with a slice of a frame of pixel data that is different from a slice of the frame of pixel data corresponding to the second stress data.
It will be understood that, although the terms “first”, “second”, “third”, etc., may be used herein to describe various elements, components, regions, layers and/or sections, these elements, components, regions, layers and/or sections should not be limited by these terms. These terms are only used to distinguish one element, component, region, layer or section from another element, component, region, layer or section. Thus, a first element, component, region, layer or section discussed herein could be termed a second element, component, region, layer or section, without departing from the spirit and scope of the inventive concept.
In addition, it will also be understood that when a layer is referred to as being “between” two layers, it can be the only layer between the two layers, or one or more intervening layers may also be present.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the inventive concept. As used herein, the terms “substantially,” “about,” and similar terms are used as terms of approximation and not as terms of degree, and are intended to account for the inherent deviations in measured or calculated values that would be recognized by those of ordinary skill in the art.
As used herein, the singular forms “a” and “an” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises” and/or “comprising”, when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. As used herein, the term “and/or” includes any and all combinations of one or more of the associated listed items. Expressions such as “at least one of,” when preceding a list of elements, modify the entire list of elements and do not modify the individual elements of the list. Further, the use of “may” when describing embodiments of the inventive concept refers to “one or more embodiments of the present disclosure”. Also, the term “exemplary” is intended to refer to an example or illustration. As used herein, the terms “use,” “using,” and “used” may be considered synonymous with the terms “utilize,” “utilizing,” and “utilized,” respectively.
It will be understood that when an element or layer is referred to as being “on”, “connected to”, “coupled to”, or “adjacent to” another element or layer, it may be directly on, connected to, coupled to, or adjacent to the other element or layer, or one or more intervening elements or layers may be present. In contrast, when an element or layer is referred to as being “directly on”, “directly connected to”, “directly coupled to”, or “immediately adjacent to” another element or layer, there are no intervening elements or layers present.
The electronic or electric devices and/or any other relevant devices or components according to embodiments of the present disclosure described herein may be implemented utilizing any suitable hardware, firmware (e.g. an application-specific integrated circuit), software, or a combination of software, firmware, and hardware. For example, the various components of these devices may be formed on one integrated circuit (IC) chip or on separate IC chips. Further, the various components of these devices may be implemented on a flexible printed circuit film, a tape carrier package (TCP), a printed circuit board (PCB), or formed on one substrate. Further, the various components of these devices may be a process or thread, running on one or more processors, in one or more computing devices, executing computer program instructions and interacting with other system components for performing the various functionalities described herein. The computer program instructions are stored in a memory which may be implemented in a computing device using a standard memory device, such as, for example, a random access memory (RAM). The computer program instructions may also be stored in other non-transitory computer readable media such as, for example, a CD-ROM, flash drive, or the like. Also, a person of skill in the art should recognize that the functionality of various computing devices may be combined or integrated into a single computing device, or the functionality of a particular computing device may be distributed across one or more other computing devices without departing from the spirit and scope of the exemplary embodiments of the present disclosure.
Although exemplary embodiments of a system and a method for mitigating the effects of compression errors have been specifically described and illustrated herein, many modifications and variations will be apparent to those skilled in the art. Accordingly, it is to be understood that to a system and a method for mitigating the effects of compression errors using stress profile compression constructed according to principles of this disclosure may be embodied other than as specifically described herein. The scope of the present disclosure is defined by the following claims, and equivalents thereof.
The present application claims priority to and the benefit of U.S. Provisional Patent Application Ser. No. 62/902,284, filed Sep. 18, 2019 and entitled “UNBIASED ITERATIVE COMPRESSION ON ADDITIVE DATA USING DITHER,” the entire content of which is hereby expressly incorporated by reference. The present application is also related to U.S. patent application Ser. No. 15/979,279, filed May 14, 2018 and entitled “STRESS PROFILE COMPRESSION,” which claims priority to and the benefit of U.S. Provisional Patent Application Ser. No. 62/643,622, filed Mar. 15, 2018 and entitled “STRESS PROFILE COMPRESSION,” the entire contents of which are hereby expressly incorporated by reference.
Number | Name | Date | Kind |
---|---|---|---|
5742892 | Chaddha | Apr 1998 | A |
6195128 | Streater | Feb 2001 | B1 |
7590299 | Brown Elliott et al. | Sep 2009 | B2 |
7911487 | Brown Elliott et al. | Mar 2011 | B2 |
8022908 | Mizukoshi et al. | Sep 2011 | B2 |
8576907 | Demos et al. | Nov 2013 | B2 |
8670625 | Sasai et al. | Mar 2014 | B2 |
8879857 | Raveendran et al. | Nov 2014 | B2 |
9392292 | Ku et al. | Jul 2016 | B2 |
9351003 | Cote et al. | Oct 2016 | B2 |
9589496 | Lu | Mar 2017 | B2 |
9799257 | Tian et al. | Oct 2017 | B2 |
20020168111 | Latva-Aho | Nov 2002 | A1 |
20050276502 | Brown Elliott | Dec 2005 | A1 |
20070229480 | Ookawara | Oct 2007 | A1 |
20120327139 | Margulis | Dec 2012 | A1 |
20140098858 | Demos et al. | Apr 2014 | A1 |
20150062202 | Lu | Mar 2015 | A1 |
20150194096 | Chung et al. | Jul 2015 | A1 |
20170347098 | Gilmutdinov et al. | Nov 2017 | A1 |
20190221158 | An | Jul 2019 | A1 |
Entry |
---|
Li, Junhui. “Image quality recovery in binary ghost imaging by adding random noise to fight quantization caused image quality decline”, “Optics Letters”, Apr. 2017, The Optical Society, vol. 42 pp. 1-8. (Year: 2017). |
Sullivan, Gary, J., et al., “Overview of the High Efficiency Video Coding (HEVC) Standard”, IEEE Transactions on Circuits and Systems for Video Technology, vol. 22, No. 12, Dec. 2012, pp. 1649-1668. |
EPO Extended Search Report dated, Apr. 15, 2019, for corresponding European Patent Application No. 19161985.7 (9 pages). |
Akyol, Emrah et al., “On Constrained Randomized Quantization,” 2012 Data Compression Conference, IEEE, Apr. 2012, pp. 72-81. |
EPO Extended Search Report dated Nov. 24, 2020, for corresponding European Patent Application No. 20188251.1 (17 pages). |
Number | Date | Country | |
---|---|---|---|
20210082341 A1 | Mar 2021 | US |
Number | Date | Country | |
---|---|---|---|
62902284 | Sep 2019 | US |