Up next in 10
AI Image Editing Qwen | Everyone can now Edit
Aug 26, 2025
AI Image editing makes photo editing accessible to anyone without any editing skills. Edit Photos with AI - https://lanhive.com
View Video Transcript
0:00
AI makes editing images very easy and
0:02
quick. You do not need any editing
0:05
skills or special software like
0:07
Photoshop. In this video, I'll show you
0:09
how we can edit images with AI using
0:12
Quen image. Let's get started. We'll
0:14
click on edit photos here. And once we
0:17
are here, we can upload the image that
0:19
we want to edit and also the settings
0:21
that we want. So for the model, I'll
0:23
choose Quinn here.
0:27
In the first one, I wanted to edit the
0:29
text. So, you can see here it says
0:31
freshly baked, open 7 a.m. to 6 p.m. So,
0:34
I wanted to change it. So, I asked when
0:37
to change the text from freshly baked to
0:39
exudas bakery. And to change the time
0:42
from what we can see here to 9 up to 7.
0:45
So, it did the exudas bakery extremely
0:47
well. You can see it maintained the font
0:49
that is in the original image. For this
0:52
one, however, the 9:00 a.m. is correct,
0:54
but instead of 5:00 p.m., it says 6:00
0:57
p.m. Overall, the editing was good. So,
1:00
for this prompt, I wanted to do two
1:02
things at once. To change the color of
1:04
the sweater and also the text that is
1:06
written here. So, you can see from the
1:08
prompt, I wanted the sweater to be
1:10
purple and the weight to say when image
1:13
edit is live on Lanhive. It did very
1:15
well overall except for the W here which
1:18
it used u instead but the rest is
1:21
extremely good. The person is maintained
1:24
exactly like in the original image. The
1:27
sweater is the same just a different
1:29
color and so on. So to be able to move
1:32
from this to this with just this prompt
1:35
is very good. It will take you a lot of
1:37
time if you are doing this let's say in
1:39
Photoshop or other editing software.
1:42
Then for this prompt, I wanted to change
1:44
three things at once. To change the
1:46
color of the sweater, to change the text
1:49
here, and also to change the background.
1:51
So you can see the prompt here. Want the
1:53
sweater to be pinged and want the words
1:55
here to say AI image editing is cool.
1:58
Then want to change the background to
2:00
the beautiful park with the blue sky.
2:02
And when you look at the edited image
2:04
here, it's beautifully done. Everything
2:07
that we requested has been done except
2:09
for the hair. You can see that the hair
2:11
there and here is slightly different. We
2:15
didn't ask it to do that. But overall,
2:17
this is very good. Again, if you're
2:19
using software like Photoshop and other
2:22
editing software, this would take a lot
2:24
of time. Then for this one, I wanted to
2:27
use this original photo to generate a
2:29
professional headsh shot for a LinkedIn
2:31
profile. So, I gave it a simple prompt.
2:34
This one perhaps this was way too simple
2:36
but you can see the results is not very
2:38
different from the first one. Just
2:40
change the biting conditions. You can
2:42
see it removed the shadows here and on
2:45
the face there. And here there was this
2:47
slight parting of the lips. You cannot
2:50
see it there. And it added this which
2:53
doesn't mean anything. So the results
2:55
were not as I expected. Maybe with good
2:57
prompting would probably get good
2:59
results. This was an interesting one. I
3:02
gave it the front view of the Nissan
3:04
Patrol V8. Then I asked it to generate
3:07
the side view. It get this result which
3:09
you can see is very impressive because
3:12
there's nothing from the original image
3:14
that is giving out this view. So this is
3:16
completely generated from the
3:18
understanding of the model. It's not
3:20
very accurate as you can see but it's
3:22
close enough. So we can see here the
3:24
side mirrors are black but here we can
3:27
see they are silver. So you can see the
3:29
outers of the windshield here black but
3:32
here it has made them white like that
3:35
and other minor issues that it
3:36
introduced but ultimately to move from
3:39
this to this without extra information
3:42
is extremely good and to be able to do
3:45
that with this simple prompt is really
3:48
really good. Then I wanted to do
3:50
something similar to the previous one
3:51
for this photo. So we had this original
3:54
photo of this family. I wanted to see
3:56
the back view if the camera was behind
3:58
them. So I gave it a very simple prompt
4:01
just said back view of the family and
4:03
this is what it gave. So you can see a
4:05
loss of details here. First of all we
4:08
can see this child in this one. Then the
4:11
man was lifting the child here. You can
4:13
see the hands are just down holding this
4:15
boy there. And the posters of the entire
4:17
family here is different from the
4:20
posters here. Then even though we placed
4:22
the camera behind them, you can see that
4:25
the view here in the background is the
4:27
same as the view here. So this isn't the
4:29
back view, but it was quite clever. And
4:32
when we look at the general features of
4:34
this family, we can also see them here,
4:36
which is not bad. For this one, I wanted
4:39
this word quail image to be behind the
4:42
red, then it's live on NHive to be in
4:46
front. So that's the prompt that I gave
4:48
it. But the results were not as I
4:51
expected. So this quen image editing
4:53
should have been behind then this should
4:56
have been in front. So it didn't do well
4:58
here. So I tried to change the prompt a
5:00
little bit but still I got the same
5:03
results. At some point I tried an
5:05
attempt similar to this. It worked very
5:07
well. For some reason it could not put
5:10
it off this time. Quen image is good
5:13
with text in English and in Chinese. So
5:15
I wanted to see if it could translate
5:17
this freshly baked to an equivalent in
5:20
Chinese. So I gave it this simple
5:22
prompt. It gave these characters which
5:24
are very convincing. But after doing the
5:26
research I realized that these
5:28
characters are not real. They're close
5:30
to the real characters but they're not
5:32
real. So they're not readable. If you
5:34
try to translate these you not get
5:36
anything but they are characters that
5:38
are close to these. So the work around
5:40
was very clever but the results are not
5:43
what we wanted. So, this was an
5:45
interesting one. I wanted to single out
5:47
a small edit in this group. So, I wanted
5:49
to change the glasses that the woman is
5:51
wearing to blue, also the top that she's
5:54
wearing to blue. So, I gave it this
5:56
prompt to change the top the lady's
5:58
wearing to blue, and also the glasses.
6:00
And it did extremely well. So, you can
6:02
see the edited image here. It changed
6:05
the grasses to blue, the top to blue,
6:08
and everything else remains the same,
6:10
which is very good. It shows that it
6:12
understands this photo. It understands
6:14
the environment and what is happening.
6:16
It knows who we are referring to and
6:18
also the specifics of our requirements.
6:21
This was extremely good. I'll say this
6:23
is one of the best edits from the test
6:26
so far. This was also an interesting
6:28
one. I asked it to change a lot of
6:30
things. You can see the prompt here is
6:32
fairly complex, but the results were
6:34
stunning. Exactly what I asked for.
6:37
Other editing airing models struggle
6:39
with such, but this was very good. Then
6:42
for this one I wanted to change the
6:43
style of the photo. So I gave it a
6:45
simple prompt to use watercolor dream
6:48
style and this was the result which is
6:50
very good. You can see everything has
6:52
been maintained just in waterc color.
6:55
Then I used the same photo to try a
6:57
different style. This time I used pix
7:00
style and the results was stunning. The
7:03
details on the edited photo are very
7:05
similar to the original just in pix
7:08
style. You can see the cross they're
7:10
wearing the environment and everything
7:12
suits the original photo. So generally I
7:15
noticed that quen image is very good for
7:17
editing images. You'll be able to do
7:20
amazing work in just a short amount of
7:22
time. If you to do these edits using
7:25
Photoshop, this will take you a lot of
7:27
time. But we can see what we're able to
7:29
do with just a few prompts here and
7:31
there. Of course, there's room for
7:33
improvement, making better prompts. But
7:35
outment you can see that this is very
7:37
attractive.
#Multimedia Software
#Online Media
#Photographic & Digital Arts