Abstract-Light field imaging has emerged as a very promising technology in the field of computational photography. Cameras are becoming commercially available for capturing real-world light fields. However, capturing high spatial resolution light fields remains technologically challenging, and the images rendered from real light fields have today a significantly lower spatial resolution compared to traditional 2D cameras. This paper describes an example-based super-resolution algorithm for light fields, which allows the increase of the spatial resolution of the different views in a consistent manner across all sub-aperture images of the light field. The algorithm learns linear projections between subspaces of reduced dimension in which reside patch-volumes extracted from the light field. The method is extended to cope with angular super-resolution, where 2D patches of intermediate sub-aperture images are approximated from neighbouring subaperture images using multivariate ridge regression. Experimental results show significant quality improvement when compared to state-of-the-art single-image super-resolution methods applied on each view separately, as well as when compared to a recent light field super-resolution techniques based on deep learning.

Sketches obtained from eyewitness descriptions of criminals have proven to be useful in apprehending criminals, particularly when there is a lack of evidence. Automated methods to identify subjects depicted in sketches have been proposed in the literature, but their performance is still unsatisfactory when using software-generated sketches and when tested using extensive galleries with a large amount of subjects. Despite the success of deep learning in several applications including face recognition, little work has been done in applying it for face photograph-sketch recognition. This is mainly a consequence of the need to ensure robust training of deep networks by using a large number of images, yet limited quantities are publicly available. Moreover, most algorithms have not been designed to operate on software-generated face composite sketches which are used by numerous law enforcement agencies worldwide. This paper aims to tackle these issues with the following contributions: 1) a very deep convolutional neural network is utilised to determine the identity of a subject in a composite sketch by comparing it to face photographs and is trained by applying transfer learning to a state-of-the-art model pretrained for face photograph recognition; 2) a 3-D morphable model is used to synthesise both photographs and sketches to augment the available training data, an approach that is shown to significantly aid performance; and 3) the UoM-SGFS database is extended to contain twice the number of subjects, now having 1200 sketches of 600 subjects. An extensive evaluation of popular and stateof-the-art algorithms is also performed due to the lack of such information in the literature, where it is demonstrated that the proposed approach comprehensively outperforms state-of-the-art methods on all publicly available composite sketch datasets.

Numerous methods that automatically identify subjects depicted in sketches as described by eyewitnesses have been implemented, but their performance often degrades when using real-world forensic sketches and extended galleries that mimic law enforcement mug-shot galleries. Moreover, little work has been done to apply deep learning for face photo-sketch recognition despite its success in numerous application domains including traditional face recognition. This is primarily due to the limited number of sketch images available, which are insufficient to robustly train large networks. This letter aims to tackle these issues with the following contributions: 1) a state-of-the-art model pre-trained for face photo recognition is tuned for face photo-sketch recognition by applying transfer learning, 2) a three-dimensional morphable model is used to synthesise new images and artificially expand the training data, allowing the network to prevent over-fitting and learn better features, 3) multiple synthetic sketches are also used in the testing stage to improve performance, and 4) fusion of the proposed method with a state-of-the-art algorithm is shown to further boost performance. An extensive evaluation of several popular and state-of-the-art algorithms is also performed using publicly available datasets, thereby serving as a benchmark for future algorithms. Compared to a leading method, the proposed framework is shown to reduce the error rate by 80.7% for viewed sketches and lowers the mean retrieval rank by 32.5% for real-world forensic sketches.

No abstract

Abstract. Identifying and apprehending suspects by matching sketches created from eyewitness and victim descriptions to mugshot photos is a slow process since law enforcement agencies lack automated methods to perform this task. This paper attempts to tackle this problem by combining Eigentransformation, a global intra-modality approach, with the Eigenpatches local intra-modality technique. These algorithms are then fused with an inter-modality method called Histogram of Averaged Orientation Gradients (HAOG). Simulation results reveal that the intraand inter-modality algorithms considered in this work provide complementary information since not only does fusion of the global and local intra-modality methods yield better performance than either of the algorithms individually, but fusion with the inter-modality approach yields further improvement to achieve retrieval rates of 94.05% at Rank-100 on 420 photo-sketch pairs. This performance is achieved at Rank-25 when filtering of the gallery using demographic information is carried out.

No abstract

No abstract

Light fields are 4D signals capturing rich information from a scene. The availability of multiple views enables scene depth estimation, that can be used to generate 3D point clouds. The constructed 3D point clouds, however, generally contain distortions and artefacts primarily caused by inaccuracies in the depth maps. This paper describes a method for noise removal in 3D point clouds constructed from light fields. While existing methods discard outliers, the proposed approach instead attempts to correct the positions of points, and thus reduce noise without removing any points, by exploiting the consistency among views in a light-field. The proposed 3D point cloud construction and denoising method exploits uncertainty measures on depth values. We also investigate the possible use of the corrected point cloud to improve the quality of the depth maps estimated from the light field.