Face detection, pose estimation, and landmark localization in the wild

By Xiangxin Zhu, Deva Ramanan. 2012

2,400 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

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

I-Pic: A Platform for Privacy-Compliant Image Capture (2016, background): “HeadHunter achieves face detection recall of ∼95% on standard image datasets like the Annotated Faces in the Wild (AFW) [12].”
Diversity in Faces (2019, neutral): “[10].”
Deep High-Resolution Representation Learning for Visual Recognition (2019, methodology): “The dataset [121] is a combination of HELEN [80], LFPW [5], AFW [193], XM2VTS and IBUG datasets, where each face has 68 landmarks.”
Latent Embeddings for Zero-Shot Classification (2016, background): “in object detection [10, 15], human pose estimation [39] and face detection [42] etc.”
DenseBox: Unifying Landmark Localization with End to End Object Detection (2015, methodology): “Zhu et al. [46] propose a tree structure model for joint face detection, pose estimation and landmark localization.”
Hierarchical Deep Temporal Models for Group Activity Recognition (2016, methodology): “HOG3D [18], spatio-temporal features [5], object interaction feature [19], and social role features [20] are used as unary component representations.”
Latent Factor Guided Convolutional Neural Networks for Age-Invariant Face Recognition (2016, methodology): “For each face image, we use the recently proposed algorithms [2, 37, 3] to detect the facial landmarks in images.”
Wing Loss for Robust Facial Landmark Localisation with Convolutional Neural Networks (2017, methodology): “The 300W dataset is a collection of multiple face datasets, including LFPW [1], HELEN [36], AFW [79] and XM2VTS [44].”
DEEP eye contact detector: Robust eye contact bid detection using convolutional neural network (2017, methodology): “1 Publicly available facial image datasets with eye contact annotations We used four facial-image datasets, LFPW [15], HELEN [10], AFW [27], and IBUG [17, 18, 19], under unconstrained conditions.”
Face Alignment in Full Pose Range: A 3D Total Solution (2015, methodology): “With 300 W, we adopt the proposed face profiling to generate 61,225 samples across large poses (1,786 from IBUG, 5,207 from AFW, 16,556 from LFPW and 37,676 from HELEN, XM2VTS is not used), which is further flipped to 122,450 samples.”
Interlinked Convolutional Neural Networks for Face Parsing (2015, background): “Model Eye Eyebrow Nose In mouth Upper lip Lower lip Mouth (all) Face Skin Overall [13] 0.”
How Far are We from Solving the 2D & 3D Face Alignment Problem? (and a Dataset of 230,000 3D Facial Landmarks) (2017, methodology): “However, as noted in [51], this error metric is biased for profile faces for which the interocular distance can be very small.”
Image and Vision Computing (2016, positive): “Note that we employ DPMs [18] to estimate the initial landmark locations for the first iteration of the above procedure.”
Max-Margin Object Detection (2015, methodology): “Moreover, our single HOG ﬁlter learned via MMOD gives a slightly better accuracy than the complex deformable part model of Zhu [22].”
Semantic part segmentation using compositional model combining shape and appearance (2014, background): “Besides, our task is part segmentation for animals of various poses and viewpoints, which appears more challenging than landmark localization for faces in [35].”
Face Alignment at 3000 FPS via Regressing Local Binary Features (2014, methodology): “Our training set consists of AFW, the training sets of LFPW, and the training sets of Helen, with 3148 images in total.”
Facial Landmark Detection by Deep Multi-task Learning (2014, methodology): “We compare against: (1) Robust Cascaded Pose Regression (RCPR) [3] using the publicly available implementation and parameter settings; (2) Tree Structured Part Model (TSPM) [32], which jointly estimates the head pose and facial landmarks; (3) A commercial software, Luxand face SDK [18]; (4) Explicit”
Facial Landmark Detection: A Literature Survey (2018, methodology): “re built for faces in different head poses (e.g. left prole, frontal, and right prole) during training. During detection, the result with the smallest tting error is considered as the nal output. In [125], multiple models are built for each discrete head pose and the best t during testing is outputt”
Face Detection without Bells and Whistles (2014, methodology): “5, the low performance DPM results are consistent with the previously reported one [36].”
Emotion Recognition in the Wild via Convolutional Neural Networks and Mapped Binary Patterns (2015, methodology): “In practice, we use the images provided by the EmotiW 2015 challenge [10], aligned using the Zhu and Ramanan facial feature detector [46].”