Datasets:

WenhaoWang
/

AnyPattern

ArXiv:

Tags:

License:

Dataset card Viewer Files Files and versions Community

Dataset Viewer

Auto-converted to Parquet

View in Dataset Viewer

Viewer

Split (1)

train · 101 rows

Pattern stringlengths 3 12	Elaboration stringlengths 25 137	Demo_1 imagewidth (px) 256 256	Demo_2 imagewidth (px) 256 256	Demo_3 imagewidth (px) 256 790	Demo_4 imagewidth (px) 256 256
Origin	The samples to add pattern.
ResizeCrop	Randomly crop and resize an image to a specified size.
Blend	Blend two images with random transparency.
GrayScale	Convert an image into grayscale.
ColorJitter	Randomly change the brightness, contrast, saturation, and hue of an image.
Blur	Randomly apply a blur filter to an image.
Pixelate	Pixelate random portions of an image.
Rotate	Randomly rotate an image within a given range of degrees.
Padding	Pad an image with random colors, height and width.
AddNoise	Add random noise to an image.
VertFlip	Flip an image vertically.
HoriFlip	Flip an image horizontally.
MemeFormat	Add a random meme onto an image.
OverlayEmoji	Overlay a random Emoji onto an image.
OverlayText	Overlay some random texts onto an image.
PerspChange	Randomly transform the perspective of an image.
OverlayImage	Randomly overlay an image onto another one.
StackImage	Randomly stack two images along the height or width dimension.
ChangeChan	Randomly shift, swap, or invert the channel of an image
EncQuality	Randomly encode (reduce) the quality of an image.
AddStripes	Randomly add different colors of stripes onto an image.
Sharpen	Randomly enhance the edge contrast of an image, making it appear more defined and detailed.
Skew	Randomly skew an image by a certain angle.
ShufPixels	Randomly rearrange (shuffle) the pixels within an image, creating a significant disruption to the original structure.
AddShapes	Randomly add shapes (such as ellipse, rectangle, triangle, star, and pentagon) onto an image.
Repeat	Randomly repeat an image for several times.
CutAssemble	Randomly cut an image into several cells and shuffle the cells.
BGChange	Randomly change the background of an image.
Xraylize	Simulate the effect of an X-ray on an image.
Mirrorize	Reflect an image along its vertical or horizontal axis.
Kaleidoscope	Replicate the effect of a kaleidoscope on an image by creating symmetrical, repeating patterns randomly.
EdgeDetect	Randomly perform edge detection by different filters.
GlassEffect	Add glass effect onto an image with random extent.
OptDistort	Simulate the effect of viewing an image through a medium that randomly distorts the light passing through it.
Weather	Add the weather effect (such as rain, fog, clouds, frost, and snow) onto an image.
SplitRotate	Randomly split and rotate an image.
Solarize	Invert all pixel values above a random threshold, creating a peculiar high-contrast effect.
Legofy	Randomly make an image look as if it is made out of 1×1 LEGO blocks.
Ink	Randomly make an image look like an ink painting.
Zooming	Randomly simulate the effect of zooming in or out while capturing an image.
CutPaste	Randomly change two parts in an image.
Cartoonize	Transforms an image into a stylized, simplified version reminiscent of a cartoon.
FDA	Apply a random Fourier Domain Adaptation to an image.
OilPaint	Randomly make an image look like an oil painting.
Jigsaw	Randomly make an image look like an Jigsaw.
FlRoCollage	A random combination of the flipped and rotated version of an image.
Elastic	Simulate a jelly-like distortion of an image.
PatchInter	Dividing two random images into patches and then interweave these patches to create a new composite image.
FancyPCA	Random alter the intensities of the RGB channels in an image, based on Principal Component Analysis (PCA) of the set of RGB pixel values.
GridDistort	Apply random non-linear distortions within a grid to an image, producing a variety of warping effects.
HistMatch	Transform an image so that its histogram matches a random histogram.
ISONoise	Apply random camera sensor noise to an image.
Multiple	A random value, drawn from a distribution, is multiplied with the pixel values of an image.
Posterize	Randomly reduce the number of bits used to represent the color components of each pixel.
Gamma	Alter the luminance values of an image by applying a power-law function.
Shadowy	Randomly add shadows onto an image.
Gravel	Randomly add gravels onto an image.
SunFlare	Randomly add sun flares onto an image.
SpeedUp	Randomly add speeding up effect onto an image.
Season	Randomly change the season of an image to spring, summer, autumn, or winter.
CorrectExpo	Correct the exposure of an image.
Mosaic	Randomly add mosaic effect on an image.
PartsCollage	Randomly collage different parts of an image.
RandCollage	Randomly collage N × M images.
Spatter	Randomly create a spatter effect on an image.
SuperPixel	SuperPixel are the random result of perceptual grouping (image oversegmentation) of pixels in an image.
Heatmap	Randomly apply a color map to an image.
PolarWarp	Randomly apply other patterns in a polar-transformed space.
DropChannel	Randomly drop a channel of an image.
ColorQuant	Randomly reduce the number of distinct colors used in an image.
Emboss	Replace each pixel of an image with a highlight or shadow to give a three-dimensional impression.
Voronoi	Randomly segment an image based on the ‘closeness’ of pixels to certain features or points.
ColorSpace	Randomly change an image with RGB space to HSV, HLS, LAB, YCrCb, LUV, XYZ, or YUV space.
OldSchool	Randomly change an image into old school style.
Ascii	Randomly use colored ASCII texts to represent an image.
PixelMelt	Randomly simulate the effect of "melting" pixels down the image.
Stylize	Randomly replace the style of an image with another one’s.
Animation	Randomly transfer an image to animation style.
AdvAttack	Randomly perform adversarial attacks to an image to generate adversarial example.
CLAHE	Apply Contrast Limited Adaptive Histogram Equalization (CLAHE) to equalize an image.
Screenshot	Emboss an image into a random screen shot.
DisLocation	Random the location of patches in an image.
DropArea	Apply DropOut or DropBlock on the image level.
WaveBlock	Apply WaveBlock on the image level.
Shape	Randomly crop an area with circle, heart, polygon, or star shape from an image.
VanGoghize	Change the style of an image to a painting from Vincent van Gogh randomly.
Moire	Add random Moire patterns onto one image.
Wavelet	Apply Wavelet to an image ans select random frequency as the final result.
Sepia	Change an image to Sepia style.
ToneCurve	Randomly change the relationship between bright and dark areas of the image by manipulating its tone curve.
Erasing	Erase a random area of an image.
AutoAug	Apply AutoAugment, which is an automated data augmentation technique discovered by the reinforcement learning, to an image.
AutoSeg	Use the Segment Anything model to preform segmentation on an image automatically.
FGRemove	Remove the foreground of an image.
ErodeDilate	Randomly add the erosion or dilation effect onto an image.
Fleet	Transform an image into one with a sense of time or the accumulation of years.
WaterWave	Add random water wave effect onto an image.
Binary	Use different methods to binarize an image.
Dehaze	Remove the haze from an image using dark channel prior.
Pyramid	Stack images to create a pyramid architecture.

End of preview. Expand in Dataset Viewer.

The dataset proposed in our paper "AnyPattern: Towards In-context Image Copy Detection".

Please go to Github for the code about how to use this dataset.

Here, we show how to download this dataset.

anypattern_v31

for letter in {a..z}; do
  wget https://huggingface.co/datasets/WenhaoWang/AnyPattern/resolve/main/train/anypattern_v31_part_a$letter
done
wget https://huggingface.co/datasets/WenhaoWang/AnyPattern/resolve/main/train/anypattern_v31_part_ba

cat anypattern_v31_part_a{a..z} anypattern_v31_part_ba > anypattern_v31.tar

tar -xvf anypattern_v31.tar

original_images

wget https://huggingface.co/datasets/WenhaoWang/AnyPattern/resolve/main/original/train_{0..19}.zip
for z in train_*.zip; do unzip $z; done
mv images/train original_images

reference_images

wget https://huggingface.co/datasets/WenhaoWang/AnyPattern/resolve/main/reference/references_{0..19}.zip
for z in references_*.zip; do unzip $z; done
mv images/references reference_images

query_k1_10_v2_test

wget https://huggingface.co/datasets/WenhaoWang/AnyPattern/resolve/main/query/query_k1_10_v2_test.tar
tar -xvf query_k1_10_v2_test.tar

query_k1_10_v2_support_select10

wget https://huggingface.co/datasets/WenhaoWang/AnyPattern/resolve/main/query/query_k1_10_v2_support_select10.tar
tar -xvf query_k1_10_v2_support_select10.tar

query_k1_10_v2_support_ori_select10

wget https://huggingface.co/datasets/WenhaoWang/AnyPattern/resolve/main/query/query_k1_10_v2_support_ori_select10.tar
tar -xvf query_k1_10_v2_support_ori_select10.tar

After downloading and unzipping all these files, we should have such a directory:

/path/to/

  anypattern_v31/
    anypattern_v31/
      0_0.jpg
      0_1.jpg
      ...

  original_images/
     T000000.jpg
     T000001.jpg
     ...

  reference_images/
     R000000.jpg
     R000001.jpg
     ...

  query_k1_10_v2_test/
     ...

  query_k1_10_v2_support_select10/
     ...

  query_k1_10_v2_support_ori_select10/
     ...

Citation

@inproceedings{
    wang2024AnyPattern,
    title={AnyPattern: Towards In-context Image Copy Detection},
    author={Wang, Wenhao and Sun, Yifan and Tan, Zhentao and Yang, Yi},
    booktitle={arXiv preprint arXiv:2404.13788},
    year={2024},
}

Contact

If you have any questions, feel free to contact Wenhao Wang (wangwenhao0716@gmail.com).

Downloads last month: 3

Edit dataset card

Size of downloaded dataset files:

25.6 MB

Size of the auto-converted Parquet files:

25.6 MB

Number of rows:

101