Auto-triage failures in A/B testing

1. A method for identifying software regressions, comprising:

obtaining first images, the first images being screenshots of user interface outputs obtained from a first version of a software component;

obtaining second images, the second images being screenshots of user interface outputs obtained from a second version of the software component;

identifying a collection of image deviations, wherein the collection of the image deviations comprises pair-wise image deviations between pairs of images,

wherein each pair of images of the pairs of images includes a first image from the first images and a corresponding second image from the second images,

wherein each image deviation comprises image data,

wherein an image deviation indicates a portion of the second image identified as differing from a spatially corresponding portion of the first image, and

wherein identifying the collection of the image de Dos comprises:

merging a first image deviation and a second image deviation in response to determining that a gap between the first image deviation and the second image deviation is less than or equal to a minimum gap size;

grouping the image deviations into deviation groups;

associating at least some of the second images with at least some of the deviation groups, wherein a second image of the at least some of the second images is associated with a deviation group in response to determining that a pair-wise image deviation identified for a pair of images that includes the second image is grouped into the deviation group; and

outputting a subset of the second images corresponding to a selected deviation group of the deviation groups responsive to a selection of an indication of the selected deviation group.

2. The method of claim 1 , wherein grouping the image deviations into the deviation groups comprises:

obtaining respective feature vectors from the image deviations; and

grouping the respective feature vectors into the deviation groups.

3. The method of claim 2 , wherein the respective feature vectors are grouped using a K-means clustering technique.

4. The method of claim 1 , wherein the portion of the second image is identified as the image deviation in response to determining that a size of the portion of the second image exceeds a threshold size.

5. The method of claim 1 , further comprising:

associating a label received from a user with one of the deviation groups.

6. The method of claim 1 , further comprising:

identifying data associated with one of the deviation groups in a ticketing software that includes artefacts related to a development process of the software component.

7. A device, comprising:

a memory; and

a processor, the processor configured to execute instructions stored in the memory to:

identify a collection of image deviations, wherein the collection of the image deviations comprises pair-wise image deviations between pairs of images,

wherein each pair of images of the pairs of images includes a first image from first images and a corresponding second image from second images,

wherein each image deviation comprises image data,

wherein an image deviation indicates a portion of the second image identified as differing from a spatially corresponding portion of the first image;

group the image deviations into deviation groups, and

wherein to identify the collection of the image deviations comprises instructions to;

merge a first image deviation and a second image deviation in response to determining that a gap between the first image deviation and the second image deviation is less than or equal to a minimum gap size;

associate at least some of the second images with at least some of the deviation groups, wherein a second image of the at least some of the second images is associated with a deviation group in response to determining that a pair-wise image deviation identified for a pair of images that includes the second image is grouped into the deviation group; and

output a subset of the second images corresponding to a selected deviation group of the deviation groups responsive to a selection of an indication of the selected deviation group.

8. The device of claim 7 , wherein the processor is further configured to execute instructions to:

perform tests on a first version of a software component to obtain the first images.

9. The device of claim 8 , wherein the processor is further configured to execute instructions to:

perform the tests on a second version of the software component to obtain the second images.

10. The device of claim 7 , wherein the instructions to group the image deviations into the deviation groups comprise instructions to:

obtain respective feature vectors from the image deviations; and

group the respective feature vectors into the deviation groups.

11. The device of claim 10 , wherein the respective feature vectors are grouped using a K-means clustering technique.

12. The device of claim 7 , wherein the portion of the second image is identified as the image deviation in response to determining that a size of the portion of the second image exceeds a threshold size.

13. The device of claim 7 , wherein the processor is further configured to execute instructions to:

associate a label received from a user with one of the deviation groups.

14. A non-transitory computer readable medium storing instructions operable to cause one or more processors to perform operations comprising:

identifying a collection of image deviations, wherein the collection of the image deviations comprises pair-wise image deviations between pairs of images,

wherein each pair of images of the pairs of images includes a first image from first images and a corresponding second image from second images,

wherein each image deviation comprises image data,

wherein an image deviation indicates a portion of the second image identified as differing from a spatially corresponding portion of the first image, and

wherein identifying the collection of the image deviations comprises:

merging a first image deviation and a second deviation in response to determining that a gap between the first image deviation and the second image deviation is less than or equal to a minimum gap size;

grouping the image deviations into deviation groups;

associating at least some of the second images with at least some of the deviation groups, wherein a second image of the at least some of the second images is associated with a deviation group in response to determining that a pair-wise image deviation identified for a pair of images that includes the second image is grouped into the deviation group; and

outputting a subset of the second images corresponding to a selected deviation group of the deviation groups responsive to a selection of an indication of the selected deviation group.

15. The non-transitory computer readable medium of claim 14 , wherein grouping the image deviations into the deviation groups comprises:

obtaining respective feature vectors from the image deviations; and

grouping the respective feature vectors into the deviation groups.

16. The non-transitory computer readable medium of claim 15 , wherein the respective feature vectors are grouped using a K-means clustering technique.

17. The non-transitory computer readable medium of claim 14 , wherein the portion of the second image is identified as the image deviation in response to determining that a size of the portion of the second image exceeds a threshold size.

18. The non-transitory computer readable medium of claim 14 , wherein the operations further comprise:

identifying data associated with one of the deviation groups in a ticketing software that includes artefacts related to a development process of a software component.

19. The non-transitory computer readable medium of claim 14 , wherein the operations further comprise:

associating a label received from a user with one of the deviation groups.

20. The non-transitory computer readable medium of claim 14 , wherein the operations further comprise:

performing tests on a first version of a software component to obtain the first images; and

performing the tests on a second version of the software component to obtain the second images.