Face alignment by coarse-to-fine shape searching

By Shizhan Zhu, Cheng Li, Chen Change Loy, Xiaoou Tang. Computer Vision and Pattern Recognition, 2015

540 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/dc661f478c5a245a919c4b3faae8ea4f598b6e64

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

ACFNet: Attentional Class Feature Network for Semantic Segmentation (2019, background): “There are a lot of successful applications of using coarse-to-fine approaches, such as face detection [13], shape detection [1], face alignment [48] and optical flow [3].”
Facial Landmark Detection: A Literature Survey (2018, methodology): “In [123], a coarse to ﬁne searching strategy is employed and the initial face shape is continuously updated based on the estimation from last stage.”
General Facial Representation Learning in a Visual-Linguistic Manner (2021, methodology): “NMEdiag ↓ NMEbox ↓ AUC 7 box ↑ Method Full Frontal Full Full CFSS [108] 3.”
How Far are We from Solving the 2D & 3D Face Alignment Problem? (and a Dataset of 230,000 3D Facial Landmarks) (2017, unknown): “Dataset 2D-FAN(Ours) MDM[40] iCCR[31] TCDCN[47] CFSS[48]”
Supervision-by-Registration: An Unsupervised Approach to Improve the Precision of Facial Landmark Detectors (2018, methodology): “We follow [39, 19, 40] to split the dataset into four sets, training, common testing, challenging testing, and full testing, respectively.”
Recombinator Networks: Learning Coarse-to-Fine Feature Aggregation (2015, methodology): “We reduce the error by 15% on the IBUG subset compared to CFSS.”
Deep High-Resolution Representation Learning for Visual Recognition (2019, methodology): “Following [149], [191], we train our models on 20 , 000 training images, and report the results on the AFLW-Full set ( 4 , 386 testing images) and the AFLW-Frontal set ( 1314 testing images selected from 4386 testing images).”
Deep Alignment Network: A Convolutional Neural Network for Robust Face Alignment (2017, methodology): “Coarse to Fine Shape Searching (CFSS) [36], similarly to SDM, uses SIFT features extracted at landmark locations.”
Neural best-buddies (2018, methodology): “ddle image in a morph sequence generated using the method of Liao et al. [2014] with different sets of sparse correspondences: (a) No ECORR term. (b) Using corresponding facial landmarks detected by [Zhu et al. 2015] (c) Using corresponding facial landmarks detected by [Kowalski et al. 2017]. featur”
Look at Boundary: A Boundary-Aware Face Alignment Algorithm (2018, methodology): “In the literature of face alignment, besides classic meth-ods (ASMs [34, 23], AAMs [13, 41, 33, 25], CLMs [29, 42] and Cascaded Regression Models [7, 5, 58, 8, 72, 73, 18]), recently, state-of-the-art performance has been achieved with Deep Convolutional Neural Networks (DC-NNs).”
Face Alignment in Full Pose Range: A 3D Total Solution (2015, methodology): “As such, fitting with model-view features is easily trapped in local minima [36].”
Mnemonic Descent Method: A Recurrent Process Applied for End-to-End Face Alignment (2016, methodology): “We do not consider the original annotations of LFPW (29-point markup) or HELEN (194point markup) as recent works [64, 37, 62] have shown that these databases have become saturated for the original an51-points 68-points”
Learning Deep Representation for Face Alignment with Auxiliary Attributes (2014, background): “it again); (7) Coarse-to-Fine Shape Searching (CFSS) [49].”
3D Face Morphable Models "In-the-Wild" (2017, background): “During the past few years, we have witnessed significant improvements in various face analysis tasks such as face detection [19, 42] and 2D facial landmark localisation on static images [40, 21, 6, 38, 43, 4, 5, 36].”
FSRNet: End-to-End Learning Face Super-Resolution with Facial Priors (2017, methodology): “It should also be noted that estimating 194 points is more difficult than 68 points [45], especially on the lowresolution faces.”
OpenFace 2.0: Facial Behavior Analysis Toolkit (2018, background): “CFSS [83] — Coarse to Fine Shape Search is a recent cascaded regression approach.”
Style Aggregated Network for Facial Landmark Detection (2018, methodology): “554 testing images from LFPW and HELEN form the common testing subset; 135 images from IBUG are regarded as the challenging testing subset.”