Webmorph will be offline for an undetermined amount of time starting Monday, January 24, 2022 in order to review potential security vulnerabilities. Don't worry, there have been no security breaches, but our IT department wants to be as cautious as possible with your personal data. You can download your files to access them during the break; Lisa will only be able to retrieve files for you in an emergency. I'm really sorry about this and hope it isn't too inconvenient.
| Access WebMorph | |
|---|---|
| We have way more demand than we can meet right now with a free service, so I am only authorising academic email addresses. However, guest login gives you access to everything that the account version does except save files between sessions. You can always download your files after each session and re-upload them. Please do not request an account unless you require long-term academic use. | |
| The following information is optional, but helps us figure out who our users are. | |
(to ID biases in account granting) |
|
| In order for this website to work properly, we have to store small files (called cookies) on your computer (one that stores your session ID to make sure only you access your account, and 3 used by Google Analytics to collect anonymous use data). Almost all websites do this, but an EU law requires that we obtain your consent first. By logging in or registering, you agree to this. | |
WebMorph is a web-based version of Psychomorph, with several additional functions. While WebMorph is optimized for averaging and transforming faces, you can delineate and average any type of image. WebMorph also has several batch functions for processing large numbers of images automatically and can even create moving gifs of your transforms. See the Help Videos under the Help menu to see some tutorials on what you can do in WebMorph.
(Double-click to demo)
Resources
 |
DeBruine, Lisa & Jones, Benedict (2017). Face Research Lab London Set. figshare. doi: 10.6084/m9.figshare.5047666.v3 |
 |
DeBruine, Lisa & Jones, Benedict (2017). Young Adult White Faces with Manipulated Versions. figshare. doi: 10.6084/m9.figshare.4220517.v1 |
 |
DeBruine, Lisa (2016). Young adult composite faces. figshare. doi: 10.6084/m9.figshare.4055130.v1 |
 |
The symmetric image scrambling methods were first published in: CA Conway, BC Jones, LM DeBruine, AC Little & A Sahraie (2008). Transient pupil constrictions to faces are sensitive to orientation and species. Journal of Vision, 8(3): 17. doi: 10.1167/8.3.17 |
 |
DeBruine, L. M. & Jones, B. C. (2015). Average Faces. Open Science Framework. osf.io/gzy7m |
 |
We are currently working on morphable KDEF composites. |
Development
WebMorph is made possible by the kind help of Bernie Tiddeman, who developed and maintains the desktop version of Psychomorph. WebMorph uses the open-source Java library FaceMorphLib and is developed and maintained by Lisa DeBruine.
WebMorph is currently in beta testing and is likely to remain so for some time. This means that there will be bugs and you cannot rely on the website being functional 100% of the time. Lisa will try to fix any problems as fast as possible, but she is the only person working on this project, so please be patient. If you're curious about the code or want to help with development, this project is open source at https://github.com/debruine/webmorph.
Webmorph will be offline for an undetermined amount of time starting Monday, January 24, 2022 in order to review potential security vulnerabilities. Don't worry, there have been no security breaches, but our IT department wants to be as cautious as possible with your personal data. You can download your files to access them during the break; Lisa will only be able to retrieve files for you in an emergency. I'm really sorry about this and hope it isn't too inconvenient.
WebMorph now has limited support for 3D faces. You can upload OBJ files and associated BMP or JPEG textures and view them in the delineator. You can also morph between two objects that have the same geometry. I will add further functions soon. New projects contain two example OBJs if you want to have a play.
The menu items have changed a bit and I've removed or changed a lot of keyboard shortcuts to make them more compatible between web browsers. See what else is new at What's New? under the Help menu.
Your project list is loading. This will take 5-30 seconds, depending on the size of your account.
| Go | - | Project | Description | Files | Owners |
|---|
Webmorph will be offline for an undetermined amount of time starting Monday, January 24, 2022 in order to review potential security vulnerabilities. Don't worry, there have been no security breaches, but our IT department wants to be as cautious as possible with your personal data. You can download your files to access them during the break; Lisa will only be able to retrieve files for you in an emergency. I'm really sorry about this and hope it isn't too inconvenient.
You can now delete folders and all of their contents in one step. Just select a single folder and cmd-delete. The full folder will go into the Trash and you can move it out by dragging. I tried to build in as many checks as possible to make sure you can't accidentally delete things you don't mean to, but file a bug report immediately if there are any unintended consequences.
Auto-Delineate uses the external company Face++, who impose usage limits on free accounts. If your auto-delineations are failing and the error message says "CONCURRENCY_LIMIT_EXCEEDED", just try later. I'm looking into how often this is used and how much a paid license with more access would cost.
File Info
File History
Drag images to reorder and double-click to remove them.
Select template points to add them to the mask. Separate points with commas, lines with semicolons, and mask areas with colons (e.g. 18,19,20,21,22 ; 22,30,29,28,18 : 23,24,25,26,27 ; 27,33,32,31,23)
Average images by selecting them in the Finder and clicking the average button, or by double-clicking on the average face to add selected images to the list. You can also drag images to the average box to add them to the list.
Recently Created Images
Transform images by selecting them in the Finder and dragging them to the 'image to transform' or 'transform dimension' boxes. You can also double-click these boxes to add the selected image.
-
image to transform
transformed image -
- transform
dimension + - % % %
- Show Continuum Settings
- to %
- to %
- to %
- (21 images in 5% steps)
- 2D Grid of Images
-
-
- Multiple Continua
- Drag images to the faces below
-
| ? | ||
| ? |
| ? | ||
| ? | ||
| ? | ||
| ? | ||
| ? | ||
| ? | ||
| ? | ||
| ? | ||
Default Alignment
| |||||||||||||
Paste your batch file from Excel into the box below.
Allowed functions in equations are:
abs(),
min(),
max(),
tan(),
sin(),
cos(),
atan(),
asin(),
acos(),
sqrt(),
pow() and
rad2deg().
X- and y-coordinates of delineation points are written, e.g.,
x[0] (the x-value of point 0). Units are in pixels.
The origin (0,0) is in the upper left corner.
- FWH
- Eye-spacing
| width | height | |
| By percent: | % | % |
| By pixel: | px | px |
Create CSV files with the Lab* values for each pixel. Check 'ignore mask' to omit pixel values that are the same as the top left pixel color.
Type positive numbers into the boxes to add margins and negative numbers to crop margins.
| 1st point: | X: | Y: | |
| 2nd point: | X: | Y: | |
| Image Size: | width: | height: |
(experimental: mask colour will be detected from the top left square if no mask is chosen)
Click on the squares you want to scramble. Drag for multiple select.
- Point Style:
- Blur:
18,19,20,21,22 ; 22,30,29,28,18)
You can set the current template as a default template so you can label the points and set up custom symmetry files. You can have as many default templates as you like. Select the current default template under Preferences (,).
Labels for the template
Paste your batch file into the box below (download template).
Color is optional and in the form of rgb(R,G,B) or transparent.
- image
- The path to the image file: e.g.,
/male/avg.jpg - align
- pt1, pt2, x1, y1, x2, y2, width, height, [color]
e.g.,0,1,497,825,853,825,1350,1800,rgb(0,0,0)orDEFAULT - resize
- width, height
e.g.,50%,50%or300px,400pxornull,400px - rotate
- degrees, [color]
e.g.,90,rgb(255,255,255) - crop
- top, right, bottom, left, [color]
e.g.,-100,100,-100,100,rgb(0,0,0) - mask
- (mask names or custom mask), blur, [color]
e.g.,(face,neck,ears),0,rgb(255,255,255)or(face),10,transparent - sym
shapeand/orcolor- mirror
trueor leave blank- order
- defaults to
align,resize,rotate,crop,mask,sym,mirror - outname
- The path to save the result to: e.g.,
/male/edited/avg.jpg
| image | align | resize | rotate | crop | mask | sym | mirror | order | outname |
|---|
Paste your batch file into the box below (download template). Shape, color and texture are in percent values (e.g., 50 or 50%).
| trans-img | from-img | to-img | shape | color | texture | outname |
|---|
Paste your batch file into the box below (download template). Put the name of each average on the first row and the images in the average in the rows below. Put each average in a new column.
| Find & replace
Separate multiple items with a semicolon (e.g., a;b;c) |
Find: Replace: |
|
| Insert prefix | ||
| Insert suffix | ||
| Add index |
This function is experimental and prone to error.
/[no prefix]**IMAGE**[no suffix].tem
Select the delineation points to delete
Modify the lines
Turn a set of images into a moving gif.
Paste in a batch file from Excel. Each column specifies one model.
- model name
- create shape model (T/F)
- create colour model (T/F)
- mask name for colour model
- list of image names
Create a PCA model for a set of templates (and optionally analyse them) or analyse a set of templates using an existing PCA model.
| .pca | |||
| .pca | |||
| .shape.csv | |||
| .pci | |||
| .pci | |||
| .color.csv | |||
| Mask for PCI |
This will create the following files:
- _analysis.shape.csv = the shape analysis of the selected templates
- _analysis.color.csv = the colour analysis of the selected images
- _model.pca = your PCA model (shape analysis)
- _model.pca.txt = a human-readable version of the PCA model
- _model.pca.var.csv = the variance information for each component in the PCA model.
- _model.pci = a directory of your PCI model (colour analysis)
- _model.tem = the procrustes normalised template for the set
- _model.jpg = the average image for the set
Reconstruct or create visualisations from an existing PCA model and
average. Paste in a tab, space or commma-delimited set of visualisations
to create. The first column should be the save path, the second column
the PCA model path, the third column the image to transform, and the
following columns are the weights for each PC. Use proportions (e.g.,
0.5 instead of 50 or 50%). For example, to create an image that is +1SD
on the first PC different from the average: /newimagename.jpg,
/_model.pca, /_average.jpg, 1.0
- T: toggle texture
- W: toggle wireframe
- S: toggle shading
- D: toggle delineation
- G: toggle grids
- C: centre object
- ← → ↑ ↓: rotate
- and ← → ↑ ↓: rotate 45°
- ← →: rotate z axis
- ← →: rotate z axis 45°
- and ← → ↑ ↓: move (small steps)
- and ↑ ↓: move z axis
- and ← → ↑ ↓: move (large steps)
- and ← → ↑ ↓: move camera view
- Drag each point to delineate the shape of the image.
- ⇧-click on a point to select or unselect it.
- -click or ⌃-click on a point to select or unselect the whole line.
- Hold shift and drag over the image select all points inside the box.
- Move a group of selected points by moving any selected point with the mouse or using the arrow keys.
- + and - to increase or decrease the size of a selected group of delineation points.
- ← and → to rotate a selected group of delineation points.
- A (or double-click on the image) to unselect all points.
- Hover over a point to see where it should go (see the footer).
- -click to add a new point.
- L to start drawing a new line. Click 'enter' to end the line.
- L to delete a line.
If the video recording is active (but not recording), activating an animation will record the frames so you can save them and create an animated gif (Tools > Moving Gif). If video recording is not active, it will play the frames at the render speed of your machine (probably faster than the duration below).
Frames per second:Duration in seconds:
WebMorph is a web-based version of Psychomorph, with several additional functions. While WebMorph is optimized for averaging and transforming faces, you can delineate and average any type of image. WebMorph also has several batch functions for processing large numbers of images automatically and can even create moving gifs of your transforms.