Topic
Upsampling
About: Upsampling is a research topic. Over the lifetime, 2426 publications have been published within this topic receiving 57613 citations.
Papers published on a yearly basis
Papers
More filters
Patent•
07 Jan 2014
TL;DR: In this article, the authors describe an encoding and decoding system for the provision of high quality digital representations of audio signals with particular attention to the correct perceptual rendering of fast transients at modest sample rates.
Abstract: Encoding and decoding systems are described for the provision of high quality digital representations of audio signals with particular attention to the correct perceptual rendering of fast transients at modest sample rates. This is achieved by optimising downsampling and upsampling filters to minimise the length of the impulse response while adequately attenuating alias products that have been found perceptually harmful.
9 citations
Posted Content•
TL;DR: Zhang et al. as mentioned in this paper developed a unified model based on a pure transformer for both RGB and RGB-D salient object detection (SOD), which takes image patches as inputs and leverages the transformer to propagate global contexts among image patches.
Abstract: Recently, massive saliency detection methods have achieved promising results by relying on CNN-based architectures. Alternatively, we rethink this task from a convolution-free sequence-to-sequence perspective and predict saliency by modeling long-range dependencies, which can not be achieved by convolution. Specifically, we develop a novel unified model based on a pure transformer, namely, Visual Saliency Transformer (VST), for both RGB and RGB-D salient object detection (SOD). It takes image patches as inputs and leverages the transformer to propagate global contexts among image patches. Apart from the traditional transformer architecture used in Vision Transformer (ViT), we leverage multi-level token fusion and propose a new token upsampling method under the transformer framework to get high-resolution detection results. We also develop a token-based multi-task decoder to simultaneously perform saliency and boundary detection by introducing task-related tokens and a novel patch-task-attention mechanism. Experimental results show that our model outperforms existing state-of-the-art results on both RGB and RGB-D SOD benchmark datasets. Most importantly, our whole framework not only provides a new perspective for the SOD field but also shows a new paradigm for transformer-based dense prediction models.
9 citations
TL;DR: Wang et al. as discussed by the authors proposed a fast and flexible convolutional neural network (FFCNN) based on DnCNN, which enjoys several desirable properties: downsampling and upscaling operations, which can sensibly reduce runtimes and memory requirements while maintaining the denoising performance.
Abstract: Seismic data denoising has always been an indispensable step in the seismic exploration workflow. The quality of the results directly affects the results of subsequent inversion and migration imaging. In this article, we proposed a fast and flexible convolutional neural network (FFCNN) based on DnCNN. In contrast to the existing DnCNN and other artificial intelligence (AI)-based denoisers, FFCNN enjoys several desirable properties: 1) downsampling and upscaling operations, which can sensibly reduce runtimes and memory requirements while maintaining the denoising performance, and 2) we introduced noised level maps, which can make that a single convolutional neural network (CNN) model is expected to inherit the flexibility of handling noise models with different parameters, even spatially variant noises by noting $M$ can be nonuniform. Another benefit of increasing the noise-level map is that we can preserve more useful seismic data information by controlling the tradeoff of noise removal effect and seismic data detail preservation. For real seismic data denoised work, the main work and advantages of this article are concentrated on the following two aspects: 1) we introduced a data augmentation strategy to overcome the lack of well-labeled samples and 2) transfer learning has been introduced to the training processing, which used the well-trained synthetic seismic data denoising network as a pretrained model. In this way, we can greatly accelerate and optimize the learning efficiency of the training network. Ultimately, we can greatly improve the computational efficiency and denoising performance based on this intelligent denoised network FFCNN. Finally, numerical experiments prove the effectiveness of our method in synthetic and real seismic data
9 citations
Patent•
09 Feb 2018
TL;DR: In this article, a single-channel sound separation method based on a convolution neural network was proposed, where the convolution layer, a pooling layer, an afusion layer, and an upsampling layer were employed.
Abstract: The invention discloses a single-channel sound separation method based on a convolution neural network, and belongs to the technical field of sound signal processing and artificial intelligence. The method comprises the steps: firstly proposing a processing frame of the single-channel sound separation method based on the convolution neural network, wherein the frame consists of the short-time Fourier transform, the convolution neural network, a time frequency mask and the inverse short-time Fourier transform, and the convolution neural network comprises a convolution layer, a pooling layer, afusion layer, and an upsampling layer. According to the invention, the characteristic that the convolution neural network is good at the mining of the spatial characteristics of two-dimensional data is employed, and the layers of a model are increased at the aspect of the number of layers of the neural network. At the aspect of the structure of the neural network, the invention proposes a convolution neural network structure which comprises a coding stage and a decoding stage. In the field of single-channel sound separation, the method, compared with a baseline model, greatly improves a separation index, and greatly reduces the number of parameters of the neural network.
9 citations
Patent•
04 Oct 2011
TL;DR: In this article, an apparatus for processing an audio signal is provided, consisting of a signal processor (110, 205, 405) and a configurator (120, 208, 408).
Abstract: An apparatus for processing an audio signal is provided. The apparatus comprises a signal processor (110; 205; 405) and a configurator (120; 208; 408). The signal processor (110; 205; 405) is adapted to receive a first audio signal frame having a first configurable number of samples of the audio signal, Moreover, the signal processor (110; 205; 405) is adapted to upsample the audio signal by a configurable upsampling factor to obtain a processed audio signal. Furthermore, the signal processor (110; 205; 405) is adapted to output a second audio signal frame having a second configurable number of samples of the processed audio signal. The configurator 120; 208; 408) is adapted to configure the signal processor (110; 205; 405) based on configuration information such that the configurable upsampling factor is equal to a first upsampling value when a first ratio of the second configurable number of samples to the first configurable number of samples has a first ratio value. Moreover, the configurator ( 120; 208; 408) is adapted to configure the signal processor (110; 205; 405) such that the configurable upsampling factor is equal to a different second upsampling value, when a different second ratio of the second configurable number of samples to the first configurable number of samples has a different second ratio value. The first or the second ratio value is not an integer value.
9 citations