BM3DCUDA - Revision history

Reel.Deal: test10

2023-01-30T22:20:59Z

test10

Reel.Deal: test5 update

2021-10-17T02:27:33Z

test5 update

Reel.Deal: Document arrays

2021-08-12T08:37:29Z

Document arrays

Reel.Deal: BM3DCUDA

2021-08-08T00:27:08Z

BM3DCUDA

New page

{{FilterCat6|External_filters|Plugins|Plugins_x64|Denoisers|Spatial-Temporal_Denoisers|Deep_color_tools}}
{{Filter3
|1=[https://github.com/WolframRhodium WolframRhodium]
|2=test3
|3=[https://github.com/WolframRhodium/VapourSynth-BM3DCUDA/issues/7 BM3DCUDA_AVS-test3.zip]
|4=Denoisers
|5=[https://www.gnu.org/licenses/gpl-2.0.txt GPLv2]
|6=[https://forum.doom9.org/showthread.php?t=183066 Doom9 Forum]
}}

== Description ==
[https://en.wikipedia.org/wiki/Block-matching_and_3D_filtering BM3D] denoising filter for [[AviSynth+]], implemented in CUDA. Also includes a cpu version implemented in AVX and AVX2 intrinsics that serves as a reference implementation on CPU. However, bitwise identical outputs are not guaranteed across CPU and CUDA implementations.
 
 
== Requirements ==
* [x86] / [x64]: [[AviSynth+]]
* Supported color formats: 32-bit Y/YUV/RGB [[planar]]
 
*CPU with [[AVX]] support ([[AVX2]] required for <code>BM3D_CPU</code>).
*CUDA-enabled GPU(s) of [https://developer.nvidia.com/cuda-gpus compute capability] 5.0 or higher (Maxwell+).
*GPU driver 450 or newer.
 

== [[Script variables|Syntax and Parameters]] ==
{{Template:FuncDef|BM3D_CPU (clip, clip "ref", float "sigma", int "block_step", int "bm_range", int "radius", int "ps_num", int "ps_range", bool "chroma")}}

{{Template:FuncDef|BM3D_CUDA (clip, clip "ref", float "sigma", int "block_step", int "bm_range", int "radius", int "ps_num", int "ps_range", bool "chroma", int "device_id", bool "fast", int "extractor_exp")}}

 
::{{Par2| |clip| }}
:::The input clip. Must be [[planar]] 32 bit float format. Each plane is denoised separately if <code>chroma</code> is set to False.
 
::{{Par2|ref|clip| }}
:::The reference clip. Must be of the same format, width, height, number of frames as clip.
 
::{{Par2|sigma|float|3.0}}
:::The strength of denoising.
:::The strength is similar (but not strictly equal) as VapourSynth-BM3D due to differences in implementation. (coefficient normalization is not implemented, for example).
 
::{{Par2|block_step|int|8.0}}
:::Sliding step to process every next reference block, valid range [1-8].
:::Total number of reference blocks to be processed can be calculated approximately by (width / block_step) * (height / block_step).
:::Smaller step results in processing more reference blocks, and is slower.
 
::{{Par2|bm_range|int|9}}
:::Length of the side of the search neighborhood for block-matching, valid range [1-8].
:::The size of search window is (bm_range * 2 + 1) x (bm_range * 2 + 1).
:::Larger is slower, with more chances to find similar patches.
 
::{{Par2|radius|int|0}}
:::The temporal radius for denoising, valid range [1, 16].
:::For each processed frame, (radius * 2 + 1) frames will be requested, and the filtering result will be returned to these frames by BM3D_VAggregate.
:::Increasing radius only increases tiny computational cost in block-matching and aggregation, and will not affect collaborative filtering, but the memory consumption can grow quadratically.
:::Thus, feel free to use large radius as long as your RAM is large enough :D
 
::{{Par2|ps_num|int|2}}
:::The number of matched locations used for predictive search, valid range [1-8].
:::Larger value increases the possibility to match more similar blocks, with tiny increasing in computational cost. But in the original MATLAB implementation of V-BM3D, it's fixed to 2 for all profiles except "lc", perhaps larger value is not always good for quality?
 
::{{Par2|ps_range|int|4}}
::: Length of the side of the search neighborhood for predictive-search block-matching, valid range [1, +inf).
 
::{{Par2|chroma|bool|false}}
:::CBM3D algorithm. Input clip must be of YUV444PS format.
:::Y channel is used in block-matching of chroma channels.
 
::{{Par2|device_id|int|0}}
:::Set GPU to be used.
 
::{{Par2|fast|bool|true}}
:::Multi-threaded copy between CPU and GPU at the expense of 4x memory consumption.
 
::{{Par2|extractor_exp|int|0}}
:::Used for deterministic (bitwise) output. This parameter is not present in the cpu version since the implementation always produces deterministic output.
:::Pre-rounding is employed for associative floating-point summation.
:::The value should be a positive integer not less than 3, and may need to be higher depending on the source video and filter parameters.
 
===BM3D_VAggregate===
BM3D_VAggregate should be called after temporal filtering.
 
 
{{Template:FuncDef|BM3D_VAggregate (clip, int "radius")}}
 
::{{Par2| |clip| }}
:::The input clip. Must be of 32 bit float format.
 
::{{Par2|radius|int|0}}
:::Same as BM3D.
 
== Examples ==
DGSource("sample.dgi")
ConvertBits(bits=32)
BM3D_CUDA(sigma=0.5, radius=2)
BM3D_VAggregate(radius=2)
ConvertBits(bits=16)
 
== Changelog ==
Version Date Changes 
test3 2021/08/06 - Separates VAggregate and compiles for AVX
test2 2021/08/01 - CPU version
test1 2021/07/25 - Initial release
 
== External Links ==
*[https://github.com/WolframRhodium/VapourSynth-BM3DCUDA/tree/avs+ GitHub] - Source code repository.
 
 
-----------------------------------------------
'''Back to [[External_filters#Spatio-Temporal_Denoisers|External Filters]] ←'''

@@ Line 2: / Line 2: @@
 {{Filter3
 |1=[https://github.com/WolframRhodium WolframRhodium]
-|2=test5 (97efb4a)
+|2=test10 (58dbc0a)
-|3=[https://github.com/WolframRhodium/VapourSynth-BM3DCUDA/issues/7 BM3DCUDA_AVS-test5.zip]
+|3=[https://github.com/WolframRhodium/VapourSynth-BM3DCUDA/issues/7 BM3DCUDA_AVS-test10.zip]
 |4=Denoisers
 |5=[https://www.gnu.org/licenses/gpl-2.0.txt GPLv2]
@@ Line 14: / Line 14: @@
 <br>
 == Requirements ==
-* [x86] / [x64]: [[AviSynth+]]
+* [x64]: [[AviSynth+]]
 * Supported color formats: 32-bit Y/YUV/RGB [[planar]]
-<br>
 *CPU with [[AVX]] support ([[AVX2]] required for <code>BM3D_CPU</code>).
 *CUDA-enabled GPU(s) of [https://developer.nvidia.com/cuda-gpus compute capability] 5.0 or higher (Maxwell+).
@@ Line 99: / Line 99: @@
 == Changelog ==
   Version      Date            Changes<br>
   test5        2021/10/16      - bm_range now defaults to 9 instead of 8, fixes parameter check
   test4        2021/09/08      - fix array parameter

@@ Line 23: / Line 23: @@
 == [[Script variables|Syntax and Parameters]] ==
-{{Template:FuncDef|BM3D_CPU (clip, clip "ref", float "sigma", int "block_step", int "bm_range", int "radius", int "ps_num", int "ps_range", bool "chroma")}}
+{{Template:FuncDef|BM3D_CPU (clip, clip "ref", float[] "sigma", int[] "block_step", int[] "bm_range", int "radius", int[] "ps_num", int[] "ps_range", bool "chroma")}}
-{{Template:FuncDef|BM3D_CUDA (clip, clip "ref", float "sigma", int "block_step", int "bm_range", int "radius", int "ps_num", int "ps_range", bool "chroma", int "device_id", bool "fast", int "extractor_exp")}}
+{{Template:FuncDef|BM3D_CUDA (clip, clip "ref", float[] "sigma", int[] "block_step", int[] "bm_range", int "radius", int[] "ps_num", int[] "ps_range", bool "chroma", int "device_id", bool "fast", int "extractor_exp")}}
 <br>
@@ Line 34: / Line 34: @@
 :::The reference clip. Must be of the same format, width, height, number of frames as clip.
 <br>
-::{{Par2|sigma|float|3.0}}
+::{{Par2|sigma|float[]|[3.0,3.0,3.0]}}
 :::The strength of denoising.
 :::The strength is similar (but not strictly equal) as VapourSynth-BM3D due to differences in implementation. (coefficient normalization is not implemented, for example).
 <br>
-::{{Par2|block_step|int|8.0}}
+::{{Par2|block_step|int[]|[8,8,8]}}
 :::Sliding step to process every next reference block, valid range [1-8].
 :::Total number of reference blocks to be processed can be calculated approximately by (width / block_step) * (height / block_step).
 :::Smaller step results in processing more reference blocks, and is slower.
 <br>
-::{{Par2|bm_range|int|9}}
+::{{Par2|bm_range|int[]|[9,9,9]}}
 :::Length of the side of the search neighborhood for block-matching, valid range [1-8].
 :::The size of search window is (bm_range * 2 + 1) x (bm_range * 2 + 1).
@@ Line 54: / Line 54: @@
 :::Thus, feel free to use large radius as long as your RAM is large enough :D
 <br>
-::{{Par2|ps_num|int|2}}
+::{{Par2|ps_num|int[]|[2,2,2]}}
 :::The number of matched locations used for predictive search, valid range [1-8].
 :::Larger value increases the possibility to match more similar blocks, with tiny increasing in computational cost. But in the original MATLAB implementation of V-BM3D, it's fixed to 2 for all profiles except "lc", perhaps larger value is not always good for quality?
 <br>
-::{{Par2|ps_range|int|4}}
+::{{Par2|ps_range|int[]|[4,4,4]}}
 ::: Length of the side of the search neighborhood for predictive-search block-matching, valid range [1, +inf).
 <br>
 ::{{Par2|chroma|bool|false}}