Heads up! To view this whole video, sign in with your Courses account or enroll in your free 7-day trial. Sign In Enroll
Preview
Start a free Courses trial
to watch this video
Learn about the Dall-E and Bing Image Creator text-to-image tools. See how they compare with a tool like Midjourney for creating design assets.
Resources
Related Discussions
Have questions about this video? Start a discussion with the community and Treehouse staff.
Sign upRelated Discussions
Have questions about this video? Start a discussion with the community and Treehouse staff.
Sign up
DALL-E or DALL-E 2, a product of OpenAI,
0:01
can translate textual descriptions into
visual renders, much like Midjourney.
0:03
Using DALL-E is straightforward.
0:08
On the DALL-E platform, users can craft
images by simply typing in a description.
0:09
Descriptions like a tranquil lakeside
morning could render several
0:14
visual interpretations, offering users
a choice in their preferred image.
0:18
Or in this case, a photograph
of an astronaut riding a horse.
0:22
As ridiculous as that may seem,
DALL-E's images certainly fit the bill.
0:26
But let's compare DALL-E's generation
abilities to Midjourney by using some of
0:29
the same prompts as before.
0:33
We'll start with the prompt,
0:35
photograph of a baker pulling
fresh cupcakes out of the oven.
0:36
Okay, DALL-E is finished and
these pictures are looking pretty good.
0:46
I think I prefer the compositions
of our Midjourney photos, but
0:49
we could alter our prompt to
dial that in more, I'm sure.
0:53
Plus I have to give it to DALL-E for
the photo realism in these images.
0:56
Outside of a couple of errors,
these are quite lifelike.
0:59
I'm gonna try generating some
variations of this cupcake image,
1:02
let's see what we get.
1:05
Not bad, but
I still like our original the most.
1:08
Since DALL-E's images have a resolution
limit of 1024 pixels by 1024 pixels,
1:11
and no way to change the aspect ratio,
1:17
we can try using DALL-E's Edit Image
feature to generate a wider photo.
1:19
Let's give it a shot.
1:25
To generate another frame,
we need to insert a prompt.
1:35
So I guess we'll just try using our
original prompt again, fingers crossed.
1:37
After a second,
DALL-E has indeed extended our image.
1:50
It has also provided us with
several options to choose from.
1:53
Not bad, not bad at all.
1:57
For the rest of our test, we'll move over
to Bing Image Creator, and let me explain
2:03
why, leveraging the groundbreaking
capabilities of DALL-E from open AI,
2:08
Bing has introduced
the Bing Image Creator.
2:12
This integration is a testament
to the collaborative partnership
2:15
between Microsoft and OpenAI,
2:18
merging the strengths of both to
offer users a transformative tool.
2:20
Bing's interface has been enhanced with
an Image Creator button in the search bar.
2:24
This gateway allows users to tap
directly into the power of DALL-E.
2:28
For those navigating via Microsoft Edge,
2:32
Bing Image Creator is conveniently
situated in the Edge sidebar.
2:35
Signing up to use the Bing Image Creator
with your Microsoft account is an easy and
2:40
cost-effective way to get started
with generative AI for free.
2:44
You start with 100 credits, with each
image you create costing one credit.
2:48
No worries if you run low,
you get 25 extra credits daily.
2:52
Even if you're out of credits,
2:56
you can still make images
with a tad more wait time.
2:58
In this first example, I'm generating
a random prompt supplied by Bing.
3:01
It chose the prompt, pixel art of a Jack
Russell Terrier floating in space with
3:05
the moon and stars an well,
that's exactly what we got.
3:09
Bing Image Creator has
done a fantastic job, but
3:13
let's try some of our prompts from before.
3:15
We'll start with a photography prompt.
3:18
Photograph of a baker pulling
fresh cupcakes out of the oven.
3:20
And just like that, the Bing Image Creator
has produced our four images.
3:30
Since Bing Image Creator
is powered by DALL-E,
3:34
it too shares the max resolution
limit of 1024 by 1024 pixels and
3:37
can only produce images in
a one-by-one aspect ratio.
3:42
Next, we'll try our background prompt.
3:46
Create a textured gradient background
pattern, subtle shades, and
3:48
hues of magenta.
3:52
Okay, not bad, but let's try another
version for a smooth gradient.
3:55
Whoa, now that's what I'm talking about.
4:09
Those are some awesome background options.
4:11
Now let's see how DALL-E handles
creating our cupcake icon.
4:25
We'll use the same prompt as earlier.
4:28
Hey, not bad at all.
4:38
I really like the first and
third images and
4:40
could easily see myself
using those in my designs.
4:42
But now let's see how DALL-E handles
generating web design inspiration.
4:48
It seems DALL-E has generated
illustrations of websites on computers,
4:58
which isn't very useful but I'm sure
we could try dialing in our prompt for
5:02
better results.
5:06
However, since we can't generate
longer portrait-style images,
5:07
we'll move on to trying our last
prompt for a mobile bakery app.
5:11
Shoot, it seems we've run into a similar
problem with the mobile app UI prompt.
5:26
Maybe we can alter it to say, mobile app
design for an online bakery instead.
5:30
I wonder if that will help
out with the results at all.
5:35
Let's try it.
5:37
Nope, we're still getting illustrations.
5:53
Now this style could work as design
elements in promo content perhaps, but
5:55
it doesn't work well for our use case.
5:59
But this will do it for our showcase
on the DALL-E image-to-text model and
6:02
the Bing Image Creator.
6:06
Let's move on to our next tool,
Adobe Firefly.
6:08
You need to sign up for Treehouse in order to download course files.
Sign upYou need to sign up for Treehouse in order to set up Workspace
Sign up