Learning to pool high-level features for face representation

Huang, Renjie; Ye, Mao; Xu, Pei; Li, Tao; Dou, Yumin

doi:10.1007/s00371-014-1049-8

Cited by 5 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different features are combined by simply concatenating them, as proposed, for instance, for pedestrian detection in Liang et al (2012). Feature pooling and/or dimensionality reduction techniques (Huang et al (2014)) might be used as well, but we prefer to stick with a simple approach and the obtained results reported in the following are promising. Similarly, we have employed an early-fusion strategy, combining the features from the very beginning, before the classification and decision take place.…”

Section: Methodsmentioning

confidence: 99%

A deep analysis on age estimation

Huerta

Fernández²,

Segura³

et al. 2015

Pattern Recognition Letters

View full text Add to dashboard Cite

The automatic estimation of age from face images is increasingly gaining attention, as it facilitates applications including advanced video surveillance, demographic statistics collection, customer profiling, or search optimization in large databases. Nevertheless, it becomes challenging to estimate age from uncontrollable environments, with insufficient and incomplete training data, dealing with strong person-specificity and high within-range variance. These difficulties have been recently addressed with complex and strongly hand-crafted descriptors, difficult to replicate and compare. This paper presents two novel approaches: first, a simple yet effective fusion of descriptors based on texture and local appearance; and second, a deep learning scheme for accurate age estimation. These methods have been evaluated under a diversity of settings, and the extensive experiments carried out on two large databases (MORPH and FRGC) demonstrate state-of-the-art results over previous work.

show abstract

Section: Methodsmentioning

confidence: 99%

A deep analysis on age estimation

Huerta

Fernández²,

Segura³

et al. 2015

Pattern Recognition Letters

View full text Add to dashboard Cite

show abstract

“…(18). 3.2 Self-Adaptive PCA dictionary learning Recently, there have been many works about learning dictionaries [14,[40][41][42][43] from natural image patches. It is well-known that the KSVD [20,21] can represent various image local structures by learning a universal dictionary from natural image dataset.…”

Section: Split Bregman Based Iterative Algorithm For the Proposed Asnmentioning

confidence: 99%

Compressed sensing image reconstruction via adaptive sparse nonlocal regularization

et al. 2016

View full text Add to dashboard Cite

Compressed Sensing (CS) has been successfully utilized by many computer vision applications. However,the task of signal reconstruction is still challenging, especially when we only have the CS measurements of an image (CS image reconstruction). Compared with the task of traditional image restoration (e.g., image denosing, debluring and inpainting, etc.), CS image reconstruction has partly structure or local features. It is difficulty to build a dictionary for CS image reconstruction from itself. Few studies have shown promising reconstruction performance since most of existing methods employed a fixed set of bases (e.g., wavelets, DCT, and gradient spaces) as the dictionary, which lack the adaptivity to fit image local structures. In this paper, we propose an adaptive sparse nonlocal regularization (ASNR) approach for CS image reconstruction. In ASNR, an effective self-adaptive learning dictionary is used to greatly reduce artifacts and the loss of fine details. The dictionary is compact and learned from the reconstructed image itself rather than natural image dataset. Furthermore, the image sparse nonlocal (or nonlocal self-similarity) priors are integrated into the regularization term, thus ASNR can effectively enhance the quality of the CS image reconstruction. To

show abstract

“…To obtain the intensity of the whole frame, the Prkachin and Solomon pain intensity (PSPI) scale was computed as the sum of intensity of AU4, the maximum intensity of AU6 and AU7, the maximum intensity of AU9 and AU10, and the one of AU10. The PSPI is in a range of [0,15]. Note that the PSPI FACS pain scale is currently the only intensity metric which measures on a frame-by-frame scheme.…”

Section: Pain Intensity Detectionmentioning

confidence: 99%

“…The second-order feature pooling takes into account the pairwise correlation of local features. Considering its superiority in performance compared with the first-order pooling method, the second-order pooling is applied to a few hot topics such as object recognition, human pose estimation [13], touch saliency prediction [14], and face representation [15], etc.…”

Section: Related Workmentioning

confidence: 99%

Capturing correlations of local features for image representation

et al. 2016

View full text Add to dashboard Cite

Local descriptors are popular ways to characterize the local properties of images in various computer vision based tasks. To form the global descriptors for the image regions, the first-order feature pooling is widely used. However, as the first-order pooling technique treats each dimension of local features separately, the pairwise correlations of local features are usually ignored.Encouraged by the success of recently developed second-order pooling techniques, in this paper we formulate a general second-order pooling framework and explore several analogues of the second-order average and max operations.We comprehensively investigate a variety of moments which are in the central positions to the second-order pooling technique. As a result, the superiority of the second-order standardized moment average pooling (2Standmap) is suggested. We successfully apply 2Standmap to four challenging tasks namely texture classification, medical image analysis, pain expression recognition, and micro-expression recognition. It illustrates the effectiveness of 2Standmap to capture multiple cues and the generalization to both static images and spatialtemporal sequences.

show abstract

Learning to pool high-level features for face representation

Cited by 5 publications

References 28 publications

A deep analysis on age estimation

A deep analysis on age estimation

Compressed sensing image reconstruction via adaptive sparse nonlocal regularization

Capturing correlations of local features for image representation

Contact Info

Product

Resources

About