Lossy video stream compression is performed to reduce the bandwidth and storage requirements. Moreover also image compression is a need that arises in many circumstances.It is often the case that older archive are stored at low resolution and with a compression rate suitable for the technology available at the time the video was created. Unfortunately, lossy compression algorithms cause artifact. Such artifacts, usually damage higher frequency details also adding noise or novel image patterns. There are several issues with this phenomenon. Low-quality images can be less pleasant to persons. Object detectors algorithms may have their performance reduced. As a result, given a perturbed version of it, we aim at removing such artifacts to recover the original image. To obtain that, one should reverse the compression process through a complicated non-linear image transformation. We propose a deep neural network able to improve image quality. We show that this model can be optimized either traditionally, directly optimizing an image similarity loss (SSIM), or using a generative adversarial approach (GAN). Our restored images have more photorealistic details with respect to traditional image enhancement networks. Our training procedure based on sub-patches is novel. Moreover, we propose novel testing protocol to evaluate restored images quantitatively. Differently from previously proposed approaches we are able to remove artifacts generated at any quality by inferring the image quality directly from data. Human evaluation and quantitative experiments in object detection show that our GAN generates images with finer consistent details and these details make a difference both for machines and humans.

Preserving low-quality video through deep learning / Galteri L.; Seidenari L.; Uricchio T.; Bertini M.; Bimbo A.D.. - ELETTRONICO. - 949:(2020), pp. 0-0. (Intervento presentato al convegno 2nd International Conference Florence Heri-Tech: The Future of Heritage Science and Technologies, HERITECH 2020 nel 2020) [10.1088/1757-899X/949/1/012068].

Preserving low-quality video through deep learning

Galteri L.;Seidenari L.;Uricchio T.;Bertini M.;Bimbo A. D.
2020

Abstract

Lossy video stream compression is performed to reduce the bandwidth and storage requirements. Moreover also image compression is a need that arises in many circumstances.It is often the case that older archive are stored at low resolution and with a compression rate suitable for the technology available at the time the video was created. Unfortunately, lossy compression algorithms cause artifact. Such artifacts, usually damage higher frequency details also adding noise or novel image patterns. There are several issues with this phenomenon. Low-quality images can be less pleasant to persons. Object detectors algorithms may have their performance reduced. As a result, given a perturbed version of it, we aim at removing such artifacts to recover the original image. To obtain that, one should reverse the compression process through a complicated non-linear image transformation. We propose a deep neural network able to improve image quality. We show that this model can be optimized either traditionally, directly optimizing an image similarity loss (SSIM), or using a generative adversarial approach (GAN). Our restored images have more photorealistic details with respect to traditional image enhancement networks. Our training procedure based on sub-patches is novel. Moreover, we propose novel testing protocol to evaluate restored images quantitatively. Differently from previously proposed approaches we are able to remove artifacts generated at any quality by inferring the image quality directly from data. Human evaluation and quantitative experiments in object detection show that our GAN generates images with finer consistent details and these details make a difference both for machines and humans.
2020
IOP Conference Series: Materials Science and Engineering
2nd International Conference Florence Heri-Tech: The Future of Heritage Science and Technologies, HERITECH 2020
2020
Galteri L.; Seidenari L.; Uricchio T.; Bertini M.; Bimbo A.D.
File in questo prodotto:
File Dimensione Formato  
Galteri_2020_IOP_Conf._Ser.%3A_Mater._Sci._Eng._949_012068.pdf

accesso aperto

Descrizione: articolo principale
Tipologia: Pdf editoriale (Version of record)
Licenza: Open Access
Dimensione 5.52 MB
Formato Adobe PDF
5.52 MB Adobe PDF

I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificatore per citare o creare un link a questa risorsa: https://hdl.handle.net/2158/1237256
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact