Papers › Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set

Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set

20 Mar 2019arXiv:1903.08527archive 2025-07-28

Yu Deng, Jiaolong Yang, Sicheng Xu, Dong Chen, Yunde Jia, Xin Tong

Recently, deep learning based 3D face reconstruction methods have shown promising results in both quality and efficiency.However, training deep neural networks typically requires a large volume of data, whereas face images with ground-truth 3D face shapes are scarce. In this paper, we propose a novel deep 3D face reconstruction approach that 1) leverages a robust, hybrid loss function for weakly-supervised learning which takes into account both low-level and perception-level information for supervision, and 2) performs multi-image face reconstruction by exploiting complementary information from different images for shape aggregation. Our method is fast, accurate, and robust to occlusion and large pose. We provide comprehensive experiments on three datasets, systematically comparing our method with fifteen recent methods and demonstrating its state-of-the-art performance.

PaperPDFCodeCode Syntology ran

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 36 code samples harvested from 3 repositories linked to this paper: 25 ran (of which 0 constructed an object rather than computing a result; 20 with no instrument failure: 3 honoured, 0 violated, 17 with no contract checked; 5 where Syntology's instrument failed) · 11 unverified. Where Syntology's instrument failed: 4 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 9 samples from 1 repository, 6 ran (6 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 27 samples from 2 repositories, 19 ran (14 with no instrument failure; 5 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 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

Microsoft/Deep3DFaceReconstruction officialmentioned in papermentioned on GitHubtfMITSyntology: 9 samples, 6 ran (0 instrument-failed) · 3 unverified report
qiuyu96/carver mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
sicxu/Deep3DFaceRecon_pytorch mentioned on GitHubpytorchMITSyntology: 9 samples, 7 ran (1 instrument-failed) · 2 unverified report
thu-ml/at3d mentioned on GitHubpytorchMITSyntology: 18 samples, 12 ran (4 instrument-failed) · 6 unverified 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

36 samples harvested: 25 ran (of which 0 constructed an object rather than computing a result; 20 with no instrument failure: 3 honoured, 0 violated, 17 with no contract checked; 5 where Syntology's instrument failed) · 11 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

3ran · honoured contract
4ran · our draft was wrong
1ran; the contract check could not run
17ran
11unverified

Licence: 1 of the 36 samples is pointer only, meaning Syntology does not serve that copy's text. For that sample, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 3 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “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.

Microsoft/Deep3DFaceReconstruction official repository 9 samples, 6 ran (0 instrument-failed) · 3 unverified
Landmark_loss losses.py ran MIT (permissive) · 4b277e9e226c1cab · report fixture: signature landmark_p:EagerTensor/3/float, landmark_label:EagerTensor/3/float · returned EagerTensor · level recorded 2026-09-25
Perceptual_loss losses.py ran MIT (permissive) · 3e3c4ecdbfe51e1d · report fixture: signature id_feature:EagerTensor/2/float, id_label:EagerTensor/2/float · returned EagerTensor · level recorded 2026-09-25
Photo_loss losses.py ran MIT (permissive) · f6e9653a3f0c6037 · report fixture: returned EagerTensor · level recorded 2026-09-25
load_img in 47 papers utils.py ran MIT (permissive) · 23dfaf89f2523bde · report fixture: signature img_path:str/0/str, lm_path:str/0/str · returned tuple · level recorded 2026-09-25
resize_n_crop_img preprocess_img.py ran MIT (permissive) · 8e794efb786fa179 · report fixture: signature img:Image/0/?, lm:ndarray/2/float, t:ndarray/1/float, s:float32/0/float, target_size:float/0/float · returned tuple · level recorded 2026-09-25
skinmask skin.py ran fingerprinted MIT (permissive) · c3e8e2c12f31272d · report fixture: signature imbgr:ndarray/3/int · returned ndarray · level recorded 2026-09-25
POS preprocess_img.py unverified MIT (permissive) · d536967c5e11597f · report
align_img preprocess_img.py unverified MIT (permissive) · 17b1774cc3712c5c · report
load_graph in 36 papers utils.py unverified MIT (permissive) · effa45a7df1ea25f · report
sicxu/Deep3DFaceRecon_pytorch community (archive-listed) 9 samples, 7 ran (1 instrument-failed) · 2 unverified
check_list util/generate_list.py ran MIT (permissive) · 9dfe076fbdaf5fff · report fixture: signature rlms_list:list/1/str, rimgs_list:list/1/str, rmsks_list:list/1/str · returned tuple · level recorded 2026-09-24
filter_state_dict in 12 papers models/networks.py ran MIT (permissive) · 0f6c5532e66c554c · report fixture: signature state_dict:dict/0/?, remove_name:str/0/str · returned dict · level recorded 2026-09-24
get_scheduler in 122 papers models/networks.py ran · our draft was wrong MIT (permissive) · 962cf09d55becbb1 · report fixture: signature optimizer:Adam/0/?, opt:BaseOptions/0/? · returned LambdaLR · level recorded 2026-09-24
load_data in 700 papers util/detect_lm68.py ran MIT (permissive) · fbc9ce8ff8e4f46e · report fixture: signature img_name:str/0/str, txt_name:str/0/str · returned tuple · level recorded 2026-09-24
perceptual_loss in 5 papers models/losses.py ran fingerprinted MIT (permissive) · 945b4307eb32c770 · report fixture: signature id_featureA:Tensor/2/float, id_featureB:Tensor/2/float · returned Tensor · level recorded 2026-09-24
perspective_projection in 7 papers models/bfm.py ran · honoured contract MIT (permissive) · 2e9ce6a02832f7c8 · report fixture: signature focal:float/0/float, center:float/0/float · returned ndarray · level recorded 2026-09-24
photo_loss models/losses.py ran MIT (permissive) · 8ee1568e301607cf · report fixture: signature imageA:Tensor/4/float, imageB:Tensor/4/float, mask:Tensor/4/float, eps:float/0/float · returned Tensor · level recorded 2026-09-24
load_lm_graph util/detect_lm68.py unverified MIT (permissive) · 5917027e9a785fa7 · report
resize_n_crop models/losses.py unverified MIT (permissive) · 2870025ec65fec68 · report
thu-ml/at3d community (archive-listed) 18 samples, 12 ran (4 instrument-failed) · 6 unverified
IOU in 17 papers AT3D/align_methods/align.py ran; the contract check could not run (RAISES) · metamorphic tier: well formed fingerprinted MIT (permissive) · 9348919b3a8be935 · report fixture: signature Reframe:list/1/float, GTframe:list/1/float · returned float · level recorded 2026-09-22
LResNet50E_IR AT3D/networks/CosFace.py ran MIT (permissive) · 32918e149a737ef6 · report fixture: signature is_gray:bool/0/bool · returned LResNet · level recorded 2026-09-25
get_block in 40 papers AT3D/networks/ArcFace.py ran · honoured contract fingerprinted MIT (permissive) · ad5dbf57f3ea2633 · report fixture: signature in_channel:int/0/int, depth:int/0/int, num_units:int/0/int, stride:int/0/int · returned list · level recorded 2026-09-22
get_blocks in 39 papers AT3D/networks/ArcFace.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · ea7941eb1ea8ffb2 · report fixture: signature num_layers:int/0/int · returned list · level recorded 2026-09-22
get_dataset_celeb AT3D/align_methods/face_image.py ran MIT (permissive) · 924034eebb04536f · report fixture: signature input_dir:str/0/str · returned list · level recorded 2026-09-25
get_dataset_webface AT3D/align_methods/face_image.py ran MIT (permissive) · baa6487e16d2b4a8 · report fixture: signature input_dir:str/0/str · returned list · level recorded 2026-09-25
l2_norm in 71 papers
census: 17 implementations from 25 papers, 8 distinct outputs across 3 buckets, one shared input per bucket this sample: bucket 1 of 3, arg 1: rank 2, kind float, dtype float32
AT3D/networks/ArcFace.py ran · our draft was wrong fingerprinted MIT (permissive) · c54fea429589425d · report fixture: signature input:Tensor/2/float, axis:int/0/int · returned Tensor · level recorded 2026-09-22
layer in 12 papers AT3D/align_methods/detect_face.py ran MIT (permissive) · fab32b621c712604 · report fixture: signature op:callable/0/? · returned function · level recorded 2026-09-25
load_property AT3D/align_methods/face_image.py ran MIT (permissive) · a0e97123727a9fb0 · report fixture: signature data_dir:str/0/str · returned EasyDict · level recorded 2026-09-25
preprocess in 284 papers
census: 11 implementations from 13 papers, 10 distinct outputs across 5 buckets, one shared input per bucket
AT3D/align_methods/face_preprocess.py ran MIT (permissive) · d467dd87dfba11d4 · report fixture: signature img:ndarray/3/int, image_size:tuple/1/int, bbox:NoneType/0/?, landmark:ndarray/2/float · returned tuple · level recorded 2026-09-24
prewhiten AT3D/networks/FaceNet.py ran · our draft was wrong fingerprinted MIT (permissive) · 4b408c0f70bc4e29 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-24
to_rgb in 6 papers AT3D/align_methods/align.py ran · honoured contract fingerprinted MIT (permissive) · 29507ad0ee182d8c · report fixture: signature img:ndarray/2/int · returned ndarray · level recorded 2026-09-22
align in 23 papers AT3D/align_methods/align.py unverified MIT (permissive) · cff9b0f0e442827c · report
create_mtcnn AT3D/align_methods/detect_face.py unverified MIT (permissive) · eb0033aab98bd766 · report
detect_face in 7 papers AT3D/align_methods/detect_face.py unverified MIT (permissive) · 96bf1c60f74fc61d · report
get_model in 368 papers AT3D/networks/FaceModel.py unverified MIT (permissive) · e1fd632cf9b9d660 · report
parse_lst_line AT3D/align_methods/face_preprocess.py unverified MIT (permissive) · 82b10a215c8cd30f · report
read_image in 73 papers AT3D/align_methods/face_preprocess.py unverified MIT (permissive) · 79625c2830b046be · report

Tasks

3D Face ReconstructionFace ReconstructionWeakly-supervised Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
3D Face Reconstruction Florence Deng RMSE Cooperative 1.60 #10 of 16 Archive leaderboard report
3D Face Reconstruction Florence Deng RMSE Indoor 1.61 #10 of 16 Archive leaderboard report
3D Face Reconstruction Florence Deep3DFaceReconstruction RMSE Cooperative 1.66 #12 of 16 Archive leaderboard report
3D Face Reconstruction Florence Deep3DFaceReconstruction RMSE Indoor 1.66 #12 of 16 Archive leaderboard report
3D Face Reconstruction Florence Deep3DFaceReconstruction RMSE Outdoor 1.69 #12 of 16 Archive leaderboard report
3D Face Reconstruction NoW Benchmark Deep3DFaceRecon PyTorch Mean Reconstruction Error (mm) 1.41 #6 of 17 Archive leaderboard report
3D Face Reconstruction NoW Benchmark Deep3DFaceRecon PyTorch Median Reconstruction Error 1.11 #6 of 17 Archive leaderboard report
3D Face Reconstruction NoW Benchmark Deep3DFaceRecon PyTorch Stdev Reconstruction Error (mm) 1.21 #6 of 17 Archive leaderboard report
3D Face Reconstruction NoW Benchmark Deng et al. 2019 Mean Reconstruction Error (mm) 1.54 #10 of 17 Archive leaderboard report
3D Face Reconstruction NoW Benchmark Deng et al. 2019 Median Reconstruction Error 1.23 #10 of 17 Archive leaderboard report
3D Face Reconstruction NoW Benchmark Deng et al. 2019 Stdev Reconstruction Error (mm) 1.29 #10 of 17 Archive leaderboard report
3D Face Reconstruction REALY Deep3D @cheek 1.528 (±0.501) #7 of 24 Archive leaderboard report
3D Face Reconstruction REALY Deep3D @forehead 2.015 (±0.449) #7 of 24 Archive leaderboard report
3D Face Reconstruction REALY Deep3D @mouth 1.368 (±0.439) #7 of 24 Archive leaderboard report
3D Face Reconstruction REALY Deep3D @nose 1.719 (±0.354) #7 of 24 Archive leaderboard report
3D Face Reconstruction REALY Deep3D all 1.657 #7 of 24 Archive leaderboard report
3D Face Reconstruction REALY (side-view) Deep3D @cheek 1.528 (±0.517) #5 of 19 Archive leaderboard report
3D Face Reconstruction REALY (side-view) Deep3D @forehead 2.074 (±0.486) #5 of 19 Archive leaderboard report
3D Face Reconstruction REALY (side-view) Deep3D @mouth 1.411 (±0.395) #5 of 19 Archive leaderboard report
3D Face Reconstruction REALY (side-view) Deep3D @nose 1.749 (±0.343) #5 of 19 Archive leaderboard report
3D Face Reconstruction REALY (side-view) Deep3D all 1.657 #5 of 19 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