神刀安全网

Colorful Image Colorization

Colorful Image Colorization
Richard Zhang
Phillip Isola
Alexei A. Efros


Colorful Image Colorization
Example input grayscale photos and output colorizations from our algorithm. These examples are cases where our model works especially well. For randomly selected examples, see the Performance comparisons section below.

Abstract


Given a grayscale photograph as input, this paper attacks the problem of hallucinating a plausible color version of the photograph. This problem is clearly underconstrained, so previous approaches have either relied on significant user interaction or resulted in desaturated colorizations. We propose a fully automatic approach that produces vibrant and realistic colorizations. We embrace the underlying uncertainty of the problem by posing it as a classification task and explore using class-rebalancing at training time to increase the diversity of colors in the result. The system is implemented as a feed-forward operation in a CNN at test time and is trained on over a million color images. We evaluate our algorithm using a "colorization Turing test", asking human subjects to choose between a generated and ground truth color image. Our method successfully fools humans 20% of the time, significantly higher than previous methods.


Our model

Colorful Image Colorization

Demo  [IPython Notebook] Caffe  [Model] [Prototxt]

Paper and Supplementary Material


Colorful Image Colorization Full paper [10MB] Colorful Image Colorization Additional details and experiments [1MB]


Performance comparisons

Click the montage to the left to see our results on Imagenet validation photos (this is an extension of Figure 6 from our paper). Click the montage to the right to see results on a test set sampled from SUN (extension of Figure 12 in our paper). These images are random samples from the test set and are not hand-selected.

Comparisons on Imagenet

Colorful Image Colorization

(hovering shows our results)

Comparisons on SUN

Colorful Image Colorization

(hovering shows our results)

We also provide an initial comparison against Cheng et al. 2015here. We were unable to acquire code or results from the authors, so we simply ran our method on screenshots from the figures in the paper of Cheng et al. See Section 3 in thesupplementary pdf for further discussion of the differences between our algorithm and that of Cheng et al.


Semantic interpretability of results

Here, we show the ImageNet categories for which our colorization helps and hurts the most on object classification. Categories are ranked according to the difference in performance of VGG classification on the colorized result compared to on the grayscale version. This is an extension of Figure 6 in the paper.

Click a category below to see our results on all test images in that category.

Top
Bottom
  1. Rapeseed

    Colorful Image Colorization

  2. Lorikeet

    Colorful Image Colorization

  3. Cheeseburger

    Colorful Image Colorization

  4. Meat Loaf

    Colorful Image Colorization

  5. Pomegranate

    Colorful Image Colorization

  1. Green Snake

    Colorful Image Colorization

  2. Pizza

    Colorful Image Colorization

  3. Yellow Lady’s Slipper

    Colorful Image Colorization

  4. Orange

    Colorful Image Colorization

  5. Goldfinch

    Colorful Image Colorization

  1. Chain

    Colorful Image Colorization

  2. Wok

    Colorful Image Colorization

  3. Can opener

    Colorful Image Colorization

  4. Water bottle

    Colorful Image Colorization

  5. Modem

    Colorful Image Colorization

  1. Standard Schnauzer

    Colorful Image Colorization

  2. Pickelhaube

    Colorful Image Colorization

  3. Half Track

    Colorful Image Colorization

  4. Barbershop

    Colorful Image Colorization

  5. Military Uniform

    Colorful Image Colorization


Acknowledgements

This research was supported, in part, by ONR MURI N000141010934, NSF SMA-1514512, and a hardware donation by NVIDIA Corp. We thank members of the Berkeley Vision Lab for helpful discussions. We thank Aditya Deshpande for providing help with comparisons to Deshpande et al.

转载本站任何文章请注明:转载至神刀安全网,谢谢神刀安全网 » Colorful Image Colorization

分享到:更多 ()

评论 抢沙发

  • 昵称 (必填)
  • 邮箱 (必填)
  • 网址
分享按钮