Burst Denoising with Kernel Prediction Networks
| Autor: | Ren Ng, Jonathan T. Barron, Dillon Sharlet, Robert J. Carroll, Ben Mildenhall, Jiawen Chen |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
FOS: Computer and information sciences
Noise measurement Computer science business.industry Computer Vision and Pattern Recognition (cs.CV) Noise reduction Computer Science - Computer Vision and Pattern Recognition ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION 020207 software engineering Pattern recognition 02 engineering and technology Convolutional neural network Maxima and minima Kernel (image processing) 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Artificial intelligence business |
| Zdroj: | CVPR |
| DOI: | 10.1109/cvpr.2018.00265 |
| Popis: | We present a technique for jointly denoising bursts of images taken from a handheld camera. In particular, we propose a convolutional neural network architecture for predicting spatially varying kernels that can both align and denoise frames, a synthetic data generation approach based on a realistic noise formation model, and an optimization guided by an annealed loss function to avoid undesirable local minima. Our model matches or outperforms the state-of-the-art across a wide range of noise levels on both real and synthetic data. To appear in CVPR 2018 (spotlight). Project page: http://people.eecs.berkeley.edu/~bmild/kpn/ |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|