Papers › Towards Flexible Blind JPEG Artifacts Removal

Towards Flexible Blind JPEG Artifacts Removal

29 Sep 2021ICCV 2021 10arXiv:2109.14573archive 2025-07-28

Jiaxi Jiang, Kai Zhang, Radu Timofte

Training a single deep blind model to handle different quality factors for JPEG image artifacts removal has been attracting considerable attention due to its convenience for practical usage. However, existing deep blind methods usually directly reconstruct the image without predicting the quality factor, thus lacking the flexibility to control the output as the non-blind methods. To remedy this problem, in this paper, we propose a flexible blind convolutional neural network, namely FBCNN, that can predict the adjustable quality factor to control the trade-off between artifacts removal and details preservation. Specifically, FBCNN decouples the quality factor from the JPEG image via a decoupler module and then embeds the predicted quality factor into the subsequent reconstructor module through a quality factor attention block for flexible control. Besides, we find existing methods are prone to fail on non-aligned double JPEG images even with only a one-pixel shift, and we thus propose a double JPEG degradation model to augment the training data. Extensive experiments on single JPEG images, more general double JPEG images, and real-world JPEG images demonstrate that our proposed FBCNN achieves favorable performance against state-of-the-art methods in terms of both quantitative metrics and visual quality.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 11 code samples harvested from 1 repository linked to this paper: 9 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 7 where Syntology's instrument failed) · 2 unverified. Where Syntology's instrument failed: 7 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 11 samples from 1 repository, 9 ran (2 with no instrument failure; 7 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

jiaxi-jiang/fbcnn officialmentioned in papermentioned on GitHubpytorchSyntology: 11 samples, 9 ran (7 instrument-failed) · 2 unverified report
olaviinha/NeuralImageSuperResolution mentioned on GitHubSyntology: not harvested report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

11 samples harvested: 9 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 7 where Syntology's instrument failed) · 2 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

7ran · our draft was wrong
2ran
2unverified

Licence: 0 of the 11 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from jiaxi-jiang/fbcnn. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

jiaxi-jiang/fbcnn official repository 11 samples, 9 ran (7 instrument-failed) · 2 unverified
QFAttention models/network_fbcnn.py ran Apache-2.0 (permissive) · 504364eda535d2ec · report fixture: signature x:Tensor/4/float, gamma:Tensor/2/float, beta:Tensor/2/float · returned Tensor · level recorded 2026-09-24
ResBlock in 127 papers
census: 6 implementations from 5 papers, 6 distinct outputs across 2 buckets, one shared input per bucket
models/network_fbcnn.py ran fingerprinted Apache-2.0 (permissive) · 2cfea441b848a2ad · report fixture: signature x:Tensor/4/float · returned Tensor · level recorded 2026-09-24
conv in 178 papers models/network_fbcnn.py ran · our draft was wrong Apache-2.0 (permissive) · 0bb3503f833a6f8e · report fixture: signature in_channels:int/0/int, out_channels:int/0/int, kernel_size:int/0/int, stride:int/0/int, padding:int/0/int, bias:bool/0/bool, mode:str/0/str, negative_slope:float/0/float · returned Sequential · level recorded 2026-09-22
downsample_avgpool models/network_fbcnn.py ran · our draft was wrong Apache-2.0 (permissive) · bf45710301b68ac5 · report fixture: signature in_channels:int/0/int, out_channels:int/0/int, kernel_size:int/0/int, stride:int/0/int, padding:int/0/int, bias:bool/0/bool, mode:str/0/str, negative_slope:float/0/float · returned Sequential · level recorded 2026-09-24
downsample_maxpool models/network_fbcnn.py ran · our draft was wrong Apache-2.0 (permissive) · b6aa67b647a44a13 · report fixture: signature in_channels:int/0/int, out_channels:int/0/int, kernel_size:int/0/int, stride:int/0/int, padding:int/0/int, bias:bool/0/bool, mode:str/0/str, negative_slope:float/0/float · returned Sequential · level recorded 2026-09-24
downsample_strideconv models/network_fbcnn.py ran · our draft was wrong Apache-2.0 (permissive) · 229a83919d75f9b5 · report fixture: signature in_channels:int/0/int, out_channels:int/0/int, kernel_size:int/0/int, stride:int/0/int, padding:int/0/int, bias:bool/0/bool, mode:str/0/str, negative_slope:float/0/float · returned Sequential · level recorded 2026-09-22
upsample_convtranspose models/network_fbcnn.py ran · our draft was wrong Apache-2.0 (permissive) · f71abbb00f56cc8a · report fixture: signature in_channels:int/0/int, out_channels:int/0/int, kernel_size:int/0/int, stride:int/0/int, padding:int/0/int, bias:bool/0/bool, mode:str/0/str, negative_slope:float/0/float · returned Sequential · level recorded 2026-09-22
upsample_pixelshuffle in 12 papers models/network_fbcnn.py ran · our draft was wrong Apache-2.0 (permissive) · 60cf0c34b7b63cb0 · report fixture: signature in_channels:int/0/int, out_channels:int/0/int, kernel_size:int/0/int, stride:int/0/int, padding:int/0/int, bias:bool/0/bool, mode:str/0/str, negative_slope:float/0/float · returned Sequential · level recorded 2026-09-22
upsample_upconv models/network_fbcnn.py ran · our draft was wrong Apache-2.0 (permissive) · 6281229c06037ca0 · report fixture: signature in_channels:int/0/int, out_channels:int/0/int, kernel_size:int/0/int, stride:int/0/int, padding:int/0/int, bias:bool/0/bool, mode:str/0/str, negative_slope:float/0/float · returned Sequential · level recorded 2026-09-22
FBCNN models/network_fbcnn.py unverified Apache-2.0 (permissive) · 9e6c6a1d38878851 · report
sequential in 6 papers models/network_fbcnn.py unverified Apache-2.0 (permissive) · 17f1c0ea54159076 · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for FBCNN (graph method: FBCNN). 11 bricks; individually on a fixture: 9 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 7 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

  1. FBCNN entry · level 0 · unverified · 9e6c6a1d38878851
  2. QFAttention model · level 1 · ran · 504364eda535d2ec
  3. ResBlock model · level 1 · ran · 2cfea441b848a2ad
  4. conv helper · level 1 · ran · our draft was wrong · 0bb3503f833a6f8e
  5. downsample_avgpool helper · level 1 · ran · our draft was wrong · bf45710301b68ac5
  6. downsample_maxpool helper · level 1 · ran · our draft was wrong · b6aa67b647a44a13
  7. downsample_strideconv helper · level 1 · ran · our draft was wrong · 229a83919d75f9b5
  8. sequential helper · level 0 · unverified · 17f1c0ea54159076
  9. upsample_convtranspose helper · level 1 · ran · our draft was wrong · f71abbb00f56cc8a
  10. upsample_pixelshuffle helper · level 1 · ran · our draft was wrong · 60cf0c34b7b63cb0
  11. upsample_upconv helper · level 1 · ran · our draft was wrong · 6281229c06037ca0

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2109.14573") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Image CompressionImage Compression Artifact ReductionImage DeblockingImage DenoisingImage ForensicsImage RestorationJPEG Artifact CorrectionJPEG Artifact Removal

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
JPEG Artifact Correction BSDS500 (Quality 10 Color) FBCNN PSNR 27.85 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 10 Color) FBCNN PSNR-B 27.52 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 10 Color) FBCNN SSIM 0.799 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 10 Grayscale) FBCNN PSNR 29.67 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 10 Grayscale) FBCNN PSNR-B 29.22 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 10 Grayscale) FBCNN SSIM 0.821 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 20 Color) FBCNN PSNR 30.14 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 20 Color) FBCNN PSNR-B 29.56 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 20 Color) FBCNN SSIM 0.867 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 20 Grayscale) FBCNN PSNR 32.00 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 20 Grayscale) FBCNN PSNR-B 31.19 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 20 Grayscale) FBCNN SSIM 0.885 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 30 Color) FBCNN PSNR 31.45 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 30 Color) FBCNN PSNR-B 30.72 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 30 Color) FBCNN SSIM 0.897 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 30 Grayscale) FBCNN PSNR 33.37 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 30 Grayscale) FBCNN PSNR-B 32.32 #1 of 2 Archive leaderboard report
JPEG Artifact Correction BSDS500 (Quality 30 Grayscale) FBCNN SSIM 0.913 #1 of 2 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 10 Grayscale) FBCNN PSNR 30.12 #2 of 7 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 10 Grayscale) FBCNN PSNR-B 29.80 #2 of 7 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 10 Grayscale) FBCNN SSIM 0.822 #2 of 7 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 20 Grayscale) FBCNN PSNR 32.31 #2 of 6 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 20 Grayscale) FBCNN PSNR-B 31.74 #2 of 6 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 20 Grayscale) FBCNN SSIM 0.872 #2 of 6 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 30 Grayscale) FBCNN PSNR 33.54 #2 of 6 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 30 Grayscale) FBCNN PSNR-B 32.78 #2 of 6 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 30 Grayscale) FBCNN SSIM 0.894 #2 of 6 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 40 Grayscale) FBCNN PSNR 34.35 #2 of 5 Archive leaderboard report
JPEG Artifact Correction Classic5 (Quality 40 Grayscale) FBCNN SSIM 0.907 #2 of 5 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 10 Color) FBCNN PSNR 32.18 #1 of 6 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 10 Color) FBCNN PSNR-B 32.15 #1 of 6 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 10 Color) FBCNN SSIM 0.815 #1 of 6 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 20 Color) FBCNN PSNR 34.38 #1 of 6 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 20 Color) FBCNN PSNR-B 34.34 #1 of 6 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 20 Color) FBCNN SSIM 0.844 #1 of 6 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 30 Color) FBCNN PSNR 35.41 #1 of 4 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 30 Color) FBCNN PSNR-B 35.35 #1 of 4 Archive leaderboard report
JPEG Artifact Correction ICB (Quality 30 Color) FBCNN SSIM 0.857 #1 of 4 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 10 Color) FBCNN PSNR 27.77 #2 of 9 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 10 Color) FBCNN PSNR-B 27.51 #2 of 9 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 10 Color) FBCNN SSIM 0.803 #2 of 9 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 20 Color) FBCNN PSNR 30.11 #2 of 9 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 20 Color) FBCNN PSNR-B 29.70 #2 of 9 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 20 Color) FBCNN SSIM 0.868 #2 of 9 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 20 Grayscale) FBCNN PSNR 32.13 #2 of 12 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 20 Grayscale) FBCNN PSNR-B 31.57 #2 of 12 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 20 Grayscale) FBCNN SSIM 0.889 #2 of 12 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 30 Color) FBCNN PSNR 31.43 #2 of 3 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 30 Color) FBCNN PSNR-B 30.92 #2 of 3 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 30 Color) FBCNN SSIM 0.897 #2 of 3 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 30 Grayscale) FBCNN PSNR 33.54 #2 of 7 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 30 Grayscale) FBCNN PSNR-B 32.83 #2 of 7 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 30 Grayscale) FBCNN SSIM 0.916 #2 of 7 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 40 Grayscale) FBCNN PSNR 34.53 #3 of 5 Archive leaderboard report
JPEG Artifact Correction LIVE1 (Quality 40 Grayscale) FBCNN SSIM 0.931 #3 of 5 Archive leaderboard report
JPEG Artifact Correction Live1 (Quality 10 Grayscale) FBCNN PSNR 29.75 #2 of 13 Archive leaderboard report
JPEG Artifact Correction Live1 (Quality 10 Grayscale) FBCNN PSNR-B 29.40 #2 of 13 Archive leaderboard report
JPEG Artifact Correction Live1 (Quality 10 Grayscale) FBCNN SSIM 0.827 #2 of 13 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections