Keywords (1)

Academic
Publications
Face Alignment Via Component-Based Discriminative Search

Face Alignment Via Component-Based Discriminative Search,10.1007/978-3-540-88688-4_6,Lin Liang,Rong Xiao,Fang Wen,Jian Sun

Face Alignment Via Component-Based Discriminative Search   (Citations: 19)
BibTex | RIS | RefWorks Download
In this paper, we propose a component-based discriminative approach for face alignment without requiring initialization 1. Unlike many approaches which locally optimize in a small range, our approach searches the face shape in a large range at the component level by a discriminative searchalgorithm. Specifically, a set of direction classifiers guide the search of the configurations of facial compo- nents among multiple detected modes of facial components. The direction clas- sifiers are learned using a large number of aligned local patches and misaligned local patches from the training data. The discriminative search is extremely effec- tive and able to find very good alignment results only in a few (2 »3) search iter- ations. As the new approach gives excellent alignment results on the commonly used datasets (e.g., AR (18), FERET (21)) created under-controlled conditions, we evaluate our approach on a more challenging dataset containing over 1,700 well-labeled facial images with a large range of variations in pose, lighting, ex- pression, and background. The experimental results show the superiority of our approach on both accuracy and efficiency.
Conference: European Conference on Computer Vision - ECCV , pp. 72-85, 2008
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ...Representative work includes Snakes [21], Active Shape Model [10] and Active Appearance Model (AAM) [8], Bayesian shape model [41], nonlinear shape models [9], [33], [42], view-based [11] and 3D models [3], [19], and models for weak initialization [25], [26], [37]...

    Yan Liet al. Robustly Aligning a Shape Model and Its Application to Car Alignment o...

    • ...2.1 Component-Based Local Features Fig. 2 shows our local feature extraction and indexing pipeline.First,fivefacialcomponents(twoeyes,nosetip,and two mouth corners) are located on a detected face [31] by a neural-network-based component detector [14]...
    • ...For each detected face, a standard facial component [14] detector is used to extract fiducial points and align the face...

    Zhong Wuet al. Scalable Face Image Retrieval with Identity-Based Quantization and Mul...

    • ...Liang et al. [9] proposed a component-based extension to this approach, where predefined components are used to guide the shape deformation with classifiers attempting to predict direction...
    • ...However, only a few methods (e.g., [4], [9], [3]) use component-landmark detectors to overcome this challenge...
    • ...In Goldmann et al. [4] and Liang et al. [9] the detector is based on Haar-wavelets which results in restricted accuracy due to their simplicity...

    Boris A. Efratyet al. Facial component-landmark detection

    • ...Recently, an algorithm based on component discriminative search is proposed for face alignment [18]...

    Yuchi Huanget al. A Component-Based Framework for Generalized Face Alignment

    • ...The vectors sampled from photos and sketches for training CITP tree are paired according to the facial landmarks detected by a state-of-the-art alignment algorithm [16]...

    Wei Zhanget al. Coupled information-theoretic encoding for face photo-sketch recogniti...

Sort by: