笔记-Handheld multi-frame super-resolution之一
1、文中反复强调的 Handheld Motion in subpixel Space 意在何处？ 1）首先，超分辨的是指由一幅低分辨率图像或一组图像序列恢复出高分辨率图像的技术。Essential ...
原文：论文速览：MFQE 2.0: A New Approach for Multi-frame Quality Enhancement on Compressed Video
MFQE . : A New Approach for Multi frame Quality Enhancement on Compressed Video 目录 MFQE . : A New Approach for Multi frame Quality Enhancement on Compressed Video . 要点 . 压缩视频特性分析 . 质量波动 . 帧间相关性 . 方法 . ...
2019-09-29 22:59 2 1033 推荐指数：
相关推荐
语音降噪论文“A Hybrid Approach for Speech Enhancement Using MoG Model and Neural Network Phoneme Classifier”的研读
最近认真的研读了这篇关于降噪的论文。它是一种利用混合模型降噪的方法，即既利用了生成模型（MoG高斯模型），也利用了判别模型（神经网络NN模型）。本文根据自己的理解对原理做了梳理。 论文是基于“Speech Enhancement Using a Mixture-Maximum Model ...
论文翻译：2020_A Hybrid DSP/Deep Learning Approach to Real-Time Full-Band Speech Enhancement
网上已经有很多人翻译了，但我做这工作只是想让自己印象更深刻 文章方向：语音增强 论文地址：基于DSP/深度学习的实时全频带语音增强方法 博客地址：https://www.cnblogs.com/LXP-Never/p/15144882.html 论文代码：https ...
论文翻译：2021_PercepNet：A Perceptually Motivated Approach for Low-complexity, Real-time Enhancement of Fullband Speech
Approach for Low-complexity, Real-time Enhancement o ...
论文笔记之：Heterogeneous Face Attribute Estimation: A Deep Multi-Task Learning Approach
Heterogeneous Face Attribute Estimation: A Deep Multi-Task Learning Approach 2017.11.28 Introduction： 人脸属性的识别在社会交互，提供了非常广泛的信息，包括 ...
论文解读（BYOL）《Bootstrap Your Own Latent A New Approach to Self-Supervised Learning》
论文标题：Bootstrap Your Own Latent A New Approach to Self-Supervised Learning 论文方向：图像领域 论文来源：NIPS2020 论文链接：https://arxiv.org/abs/2006.07733 论文代码 ...
[论文理解] Bootstrap Your Own Latent A New Approach to Self-Supervised Learning
Bootstrap Your Own Latent A New Approach to Self-Supervised Learning Intro 文章提出一种不需要负样本来做自监督学习的方法，提出交替更新假说解释EMA方式更新target network防止collapse的原因 ...
Multi-Agent Reinforcement Learning Based Frame Sampling for Effective Untrimmed Video Recognition
Multi-Agent Reinforcement Learning Based Frame Sampling for Effective Untrimmed Video Recognition ICCV 2019 (oral) 2019-08-01 15:08:19 Paper ...