CANOVA — visual attention software for designers and photographers
Made in Italy
The newparadigmin visual attentionsoftware forphotographers and designers
Cognitive Attention · Neural Observation of Visual Artwork
CANOVA is the first commercial visual-attention neural network with maps separated in time, up to 7s
Attention over time
Legibility
Automatic contrast reading for legibility that is always optimal, against the regulatory thresholds
Colour blindness
Automatic legibility analysis for colour-vision deficiencies
8% of men cannot see the colours you design
and no visual saliency software shows it to you
Don’t lose audience: design for everyone!
Photography
The first to offer attention maps in real time, via tethering, while you shoot.
Software made for the shoot: move a light, change a pose and the map updates with every shot
Upload the images, choose the benchmarksMaps and reports arrive on their own
Over 320 million parameters
Trained on more than 1 million images
and thousands of eye-tracking sessions from real people
ApplicationsWherever an image has to be looked at
Plugins
Maps from 1 s to 7 s as layers, text and contrast analysis, colour blindness and much more
Basic features are also compatible with Lightroom Classic
Many more coming soon!
Visual attention
Not just where the eye falls
but when
CANOVA is the first commercial neural architecture to produce
maps separated in time from as early as 1 s and up to 7 s
A single map can never tell you what the eye sees
at the moment that matters
1 s
3 s
7 sWhy CANOVA?
Because a single map will never show you all the information
the different kinds of audience and the different uses
0 s
Most competitors
1 single map in the first 3–5 s
On-device
The first that runs on your Mac
forever
CANOVA runs comfortably on-deviceEven on inexpensive M1 MacBook Airs with just 8GB of RAM
No credits, no tokens: use it as much as you like
Also available as a lifetime licence
Photography
Not just graphics
also photography
CANOVA has a model made specifically for photographers
Pick the most memorable shots, check which crops work best and see where the eye will fall before going to print
Time
Three maps
not one
Each map is true for one precise moment, and no single moment tells the whole image
- 1 sYou only see what imposes itself
- 3 sIf the eye agrees to settle
- 7 sFound only by searching
CANOVA computes all three and sets them side by side
It does not hold. In the first second attention is split between brand and product. Then the brand stops being interesting
Within the first 3s attention collapses onto the product and never leaves it
In the first 7s the attention gathered by the claim is too little to fix it in memoryA forgotten claim is not a claim
The lower part stays cold throughoutspace that does not perform is a missed opportunity
Where the eye was at 1s and no longer is at 7s
Does attention stay on the key elements, or slip elsewhere?
Where attention moves to and stays over time
Does the artwork lead to the CTA, or does the eye give up first?
The only software that measures the shift of attention over time
Was the goal to let the eye travel, or to focus it on a few elements?
