Media content forgery is widely spread over the Internet and has raised severe societal concerns. With the development of deep learning, new technologies such as generative adversarial networks (GANs) and media forgery technology have already been utilized for politicians and celebrity forgery, which has a terrible impact on society. Existing GAN-generated face detection approaches rely on detecting image artifacts and the generated traces. However, these methods are model-specific, and the performance is deteriorated when faced with more complicated methods. What’s more, it is challenging to identify forgery images with perturbations such as JPEG compression, gamma correction, and other disturbances. In this paper, we propose a global–local facial fusion network, namely GLFNet, to fully exploit the local physiological and global receptive features. Specifically, GLFNet consists of two branches, i.e., the local region detection branch and the global detection branch. The former branch detects the forged traces from the facial parts, such as the iris and pupils. The latter branch adopts a residual connection to distinguish real images from fake ones. GLFNet obtains forged traces through various ways by combining physiological characteristics with deep learning. The method is stable with physiological properties when learning the deep learning features. As a result, it is more robust than the single-class detection methods. Experimental results on two benchmarks have demonstrated superiority and generalization compared with other methods.

Nowadays, deepfake detection on subtle-expression manipulation, facial-detail modification, and smeared images has become a research hotspot. Existing deepfake-detection methods on the whole face are coarse-grained, where the details are missing due to the negligible manipulated size of the image. To address the problems, we propose to build a transformer model for a deepfake-detection method by organ, to obtain the deepfake features. We reduce the detection weight of defaced or unclear organs to prioritize the detection of clear and intact organs. Meanwhile, to simulate the real-world environment, we build a Facial Organ Forgery Detection Test Dataset (FOFDTD), which includes the images of mask face, sunglasses face, and undecorated face collected from the network. Experimental results on four benchmarks, i.e., FF++, DFD, DFDC-P, Celeb-DF, and for FOFDTD datasets, demonstrated the effectiveness of our proposed method.

The properties and limitations of human visual system (HVS) are usually incorporated into objective video quality techniques for better correlation with subjective methods. In this paper, we propose a new objective assessment algorithm based on reduced reference video quality metrics (RR-VQM). The key point of our method is to design multilayer Spatial-Temporal (ST) regions based on human's fovea region, which can utilize the nonuniform resolution properties of HVS for high definition (HD) videos. Quality features are then extracted from ST regions for computing the video quality. Experimental results shows that this kind of multi-layer ST regions can improve the performance of objective methods, not only the results of evaluation, but can also accelerate the evaluation speed and reduce the amount of quality parameters to a large extent.

Albeit high dynamic range (HDR) images contain way richer information than low dynamic range (LDR) images, they need to be tone mapped for visualization on traditional display devices. The community has developed diverse tone-mapping operators (TMOs), and they produce images with varying qualities. Therefore, developing effective evaluation methods consistent with human visual perception is an urgent necessity for determining the best TMO in specific scenarios, or accordingly optimizing the parameters of a certain TMO. Towards this end, we propose a new blind quality assessment model by simulating the mechanism of human visual processing (HVS), which can adaptively adjust the sensitivity according to the chromatic properties of scenes at the beginning and well represent the perception process. Specifically, we first obtain all retinal response maps of four visual sensitivities from three opponent color channels. After that, gradient similarities in each color channel are calculated as global features to represent the procedure of visual experience, and local mean values and standard deviations are computed in the brightest and darkest regions of the maps to measure the distortions caused by over- and under-exposure. Meanwhile, the maps’ natural scene statistics (NSS) are utilized as the third set of features. Finally, all these features are blended for quality assessment by support vector regression (SVR). Extensive experiments on two public benchmark databases show our method correlates highly with subjective perception and outperforms the other state-of-the-art quality assessment methods.

ITM (inverse tone-mapping) converts SDR (standard dynamic range) footage to HDR/WCG (high dynamic range /wide color gamut) for media production. It happens not only when remastering legacy SDR footage in front-end content provider, but also adapting on-the-air SDR service on user-end HDR display. The latter requires more efficiency, thus the pre-calculated LUT (look-up table) has become a popular solution. Yet, conventional fixed LUT lacks adaptability, so we learn from research community and combine it with AI. Meanwhile, higher-bit-depth HDR/WCG requires larger LUT than SDR, so we consult traditional ITM for an efficiency-performance trade-off: We use 3 smaller LUTs, each has a non-uniform packing (precision) respectively denser in dark, middle and bright luma range. In this case, their results will have less error only in their own range, so we use a contribution map to combine their best parts to final result. With the guidance of this map, the elements (content) of 3 LUTs will also be redistributed during training. We conduct ablation studies to verify method’s effectiveness, and subjective and objective experiments to show its practicability. Code is available at: https://github.com/AndreGuo/ITMLUT.