We present a no-reference blur metric for images and video. The is based on the analysis of spread edges in an image. Its perceptual significance validated through subjective experiments. novel near real-time, has low computational complexity shown to perform well over range image content. Potential applications include optimization source coding, network resource management autofocus capturing device.

Efficient point cloud compression is fundamental to enable the deployment of virtual and mixed reality applications, since number points code can range in order millions. In this paper, we present a novel data-driven geometry method for static clouds based on learned convolutional transforms uniform quantization. We perform joint optimization both rate distortion using trade-off parameter. addition, cast decoding process as binary classification occupancy map. Our outperforms MPEG reference...

In this paper, we address the problem of privacy protection in video surveillance. We introduce two efficient approaches to conceal regions interest (ROIs) based on transform-domain or codestream-domain scrambling. first technique, sign selected transform coefficients is pseudorandomly flipped during encoding. second method, some bits codestream are inverted. more specifically cases MPEG-4 as it today prevailing standard surveillance equipment. Simulations show that both techniques...

JPEG XR is the newest image coding standard from committee. It primarily targets representation of continuous-tone still images such as photographic and achieves high quality, on par with 2000, while requiring low computational resources storage capacity. Moreover, it effectively addresses needs emerging dynamic range imagery applications by including support for a wide formats.

In this paper we describe a database containing subjective assessment scores relative to 78 video streams encoded with H.264/AVC and corrupted by simulating the transmission over error-prone network. The data has been collected from 40 subjects at premises of two academic institutions. Our goal is provide balanced comprehensive enable reproducible research results in field quality assessment. order support works on full-reference, reduced-reference no-reference algorithms, both uncompressed...

A computationally fast tone mapping operator (TMO) that can quickly adapt to a wide spectrum of high dynamic range (HDR) content is quintessential for visualization on varied low (LDR) output devices such as movie screens or standard displays. Existing TMOs successfully tone-map only limited number HDR and require an extensive parameter tuning yield the best subjective-quality tone-mapped output. In this paper, we address problem by proposing fast, parameter-free scene-adaptable deep...

Point clouds have been recognized as a crucial data structure for 3D content and are essential in number of applications such virtual mixed reality, autonomous driving, cultural heritage, etc. In this paper, we propose set contributions to improve deep point cloud compression, i.e.: using scale hyperprior model entropy coding; employing deeper transforms; different balancing weight the focal loss; optimal thresholding decoding; sequential training. addition, present an extensive ablation...

In this paper, we address the problem of scrambling regions interest in a video sequence for purpose preserving privacy surveillance. We propose an efficient solution based on transform-domain scrambling. More specifically, sign selected transform coefficients is pseudo-randomly flipped during encoding. more specifically two cases MPEG-4 and Motion JPEG 2000. Simulation results show that technique can be successfully applied to conceal information scene while providing with good level...

Rate-distortion optimization (RDO) is widely used in video coding, which plays a critical role enhancing the coding efficiency substantially. Currently, RDO process performed way that of each unit (CU) maximized independently without considering dependency among CUs. As we know, current hybrid structure, spatial/temporal prediction techniques are extensively used, introduce strong In this paper, investigate with inter-frame dependency, where impact performance CU on following frames...

This paper proposes an extended selective encryption (SE) method for both H.264/advanced video coding (AVC) (CABAC) and High Efficiency Video Coding (HEVC) streams, addressing the main security issue that SE is facing: content protection, related to amount of information leakage through a protected video. Our contribution improvement in visual distortion induced by approaches. Previous works on H.264/AVC HEVC limit bins treated one specific mode CABAC-its bypass mode-which has advantage...

Point clouds are becoming essential in key applications with advances capture technologies leading to large volumes of data. Compression is thus for storage and transmission. In this work, the state art geometry attribute compression methods a focus on deep learning based approaches reviewed. The challenges faced when compressing attributes considered, an analysis current address them, their limitations relations between traditional ones. Current open questions point cloud compression,...

The issue of privacy protection in video surveillance has drawn a lot interest lately. However, thorough performance analysis and validation is still lacking, especially regarding the fulfillment privacy-related requirements. In this paper, we put forward framework to assess capacity solutions hide distinguishing facial information conceal identity. We then conduct rigorous experiments evaluate face recognition algorithms applied images altered by techniques. Results show ineffectiveness...

Existing techniques to compress point cloud attributes leverage either geometric or video-based compression tools. We explore a radically different approach inspired by recent advances in representation learning. Point clouds can be interpreted as 2D manifolds 3D space. Specifically, we fold grid onto and map from the folded using novel optimized mapping method. This results an image, which opens way apply existing image processing on attributes. However, this process is lossy nature,...

In this paper, we introduce a scheme for coding video surveillance camera networks. It is based on multiview and Distributed Video Coding (DVC). DVC known its low complexity encoders. This makes it very interesting wide range of applications, in particular surveillance. More specifically, new fusion technique between temporal side information homography-based that improves the rate-distortion performance compression. Finally, with pure Wyner-Ziv camera, which encodes all frames as frames.

In this paper, we address the problem of privacy in video surveillance systems. More specifically, consider case H.264/AVC which is state-of-the-art coding. We assume that regions interest (ROI), containing privacy-sensitive information, have been identified. The content these are then concealed using scrambling. introduce two region-based scrambling techniques. first one pseudo-randomly flips sign transform coefficients during encoding. second performing a pseudo-random permutation block....

Due to the much larger luminance and contrast characteristics of high dynamic range (HDR) images, well-known objective quality metrics, widely used for assessment low (LDR) content, cannot be directly applied HDR images in order predict their perceptual fidelity. To overcome this limitation, advanced fidelity such as HDR-VDP, have been proposed accurately visually significant differences. However, complex calibration may make them difficult use practice. A simpler approach consists computing...