We design Gemino, a new neural compression system for video conferencing based on a novel high-frequency-conditional super-resolution pipeline. Gemino upsamples a very low-resolution version of each target frame
deep neural networkimage reconstructionlow measurement ratesCompressed sensing (CS) is a popular data compression theory for many computer vision tasks, but the high reconstruction complexity for images prevents it from being used in many real-world applications. Existing end-to-end learning methods ...
a new neural compression system for video conferencing based on a novel high-frequency-conditional super-resolution pipeline. Gemino upsamples a very low-resolution version of each target frame while enhancing high-frequency details (e.g., skin texture, hair, etc.) based on information extracted ...
Image hashing is an efficient technology for processing digital images and has been successfully used in image copy detection, image retrieval, image authentication, image quality assessment, and so on. In this paper, we design a new image hashing with c
Robust digital image watermarking using quantization and back propagation neural network. Contemp. Eng. Sci. 5, 137-147.Nallagarla Ramamurthy and S.Varadrrajan, " Robust Digital Image Watermarking using Quantization and Back propagation Neural Network" ,Contemporary Engineering Sciences, Vol.5,2012,...
Multimodal medical image segmentation using multi-scale context-aware network 2022, Neurocomputing Show abstract Deep neural network compression through interpretability-based filter pruning 2021, Pattern Recognition Show abstract This paper proposes a method to compress deep neural networks (DNNs) based on...
The theory of sparse signal representations has been used to develop state of the art single image super resolution algorithms [8]. Recently, deep convolutional neural networks (CNN) have been successfully deployed for both image denoising and super resolution [9]. It is shown that state of the...
29 develop a smartphone application for herb image classification, but the entire process is run on the cloud server. The above studies have used the Deep Neural Network (DNN), which combines with powerful and expensive computation hardware to support the computation. However, due to the high ...
Specifically, the attack is simulated by multiplying with a 0–1 mask of the same size as the encoded image. In this mask, the region with pixel value 0 represents the cropped region, while the region with pixel value 1 represents the remaining region. 2.4.3. JPEG Compression The steps ...
2.1 JPEG recompression There usually are two processes in JPEG recompression: spatial image restoration and JPEG image compression. Suppose an 8 × 8 DCT block before JPEG recompres- sion is Do , and the quality factor of the image is qo , which corresponds to the quantization ...