Upload a full-body photo, design the squat you finish in, then animate from standing to that pose and add the fuzzy dream caption.
Last checked 2026-11-02
⚙ AI Forge Studio
Design the pose you finish in from your full-body photo, then animate from standing to it.
1Still · 3→
2Animate 10 s · 280
Your photo
9:16 vertical · your photo is the first frame, the designed frame is the last
Next: animate 10 s · 280 credits.
Joining the trend is free; making it on Lovart ME uses credits from packs starting at $9.90 for 250 credits. Each button shows its exact cost before you generate.
SAMPLE · INPUT → RESULT
Sample inputs and result from this page. Your result will appear here.
Start with an idea
01 / Samples
Fuzzy dream dance frame samples
Each sample shows the photo, the designed finish pose, the first and last frame and the clip, remixed from aiaigenerator.ai originals with original adult characters.
The fuzzy dream dance is the TikTok groove set to “I just woke up from a fuzzy dream”: a loose standing bounce, a drop into a deep wide squat, arm winds, both hands on the head and a fast knee shake. The people below are original generated adults, not real dancers. Both clips are Lovart ME re-cuts of 10-second Seedance 2.0 Fast originals made on aiaigenerator.ai, reframed and graded, not mirrored and with no new generation. The pose reference, the two finish poses and the caption card were made with GPT Image 2.5 Flare for this page on November 2, 2026. None came from the workbench above.
Stand to squat
Photo →Still→StartEnd
Stand to squat
Remixed from an aiaigenerator.ai clip · Seedance 2.0 Fast
Photo
GPT Image 2.5 Flare: an original generated woman with curly dark hair, a grey tee and striped pyjama trousers, barefoot, head to toes
First frame
GPT Image 2.5 Flare image to image, 9:16: the same woman standing in a sunny bedroom
Master
Seedance 2.0 Fast · 720p · 9:16 · 10 s from the first frame, made on aiaigenerator.ai
This cut
8.4 s re-cut for Lovart ME: reframed, ice-and-brass grade, not mirrored; zero credits
What happens
Hands over her eyes as she wakes, a drop into the deep squat by 2 s, arm winds, hands on her head, fists at her chest, then she stands with one arm up
Remixed from an aiaigenerator.ai clip · Seedance 2.0 Fast
Photos
Two GPT Image 2.5 Flare full-body photos of original adults: a man in a green tracksuit, a woman in a white crop top and black cargo trousers
First frame
GPT Image 2.5 Flare image to image, 9:16: both standing on a backyard lawn in front of a brick wall
Master
Seedance 2.0 Fast · 720p · 9:16 · 10 s from the first frame, made on aiaigenerator.ai
This cut
8.8 s re-cut for Lovart ME, steel-and-frost grade, not mirrored; zero credits
What happens
Both drop into the squat within 1.5 s, bounce low with fists at the chest, put both hands on their heads at about 6 s, then stand and he throws an arm up
The standing first frame from Stand to squat, and a pose reference: a wooden artist mannequin in a deep wide squat with both hands on its head (GPT Image 2.5 Flare)
Still
GPT Image 2.5 Flare image to image, 1024 × 1536, made November 2, 2026: the same woman, room and clothes in that squat, hands on her head
Add your full-body photo and one pose reference, then ask for you in a deep wide squat with both hands on your head, same face and clothes.
Your photo
One adult standing, head to toes in view, good light, no sunglasses or hat over the face. Wide trousers or shorts move best. This photo is also the first frame of the clip, so use a vertical one.
The pose reference
Optional: one picture of the pose, such as a mannequin, a sketch or a still of yourself. Only the body position is copied, never a face or clothes. Without one, the chip you pick sets the pose.
Keep your face and clothes
Design the pose sends both to Image to Image and asks for the same face, hair, clothes, room and camera with only the pose changed. Compare the two frames; if anything else moved, press Rebuild last frame.
Squat Finish
A deep wide squat, knees out, both hands on top of the head. The pose the trend ends on, and the default.
Set I just woke up from a fuzzy dream across the top in one line, so it never covers your knees or hands.
Still only · GPT Image 2.5 Flare
The one line
“I just woke up from a fuzzy dream” is the opening line of Madonna's “Miles Away” (album Hard Candy, 2008). The sound most clips use is a sped-up dance remix posted on TikTok on January 9, 2025 as an original sound by DjLukeyC; it was in 13,900 videos by October 29, 2026. Use the line on its own, in one of these forms:
I just woke up from a fuzzy dream
I JUST WOKE UP FROM A FUZZY DREAM
i just woke up from a fuzzy dream…
Where it sits
Across the top, above your head, in plain white type on a thin dark band. The knees and hands carry the dance, so keep the middle and bottom of the frame clear. Download frame, then Open in PhotoCraft and type the line. PhotoCraft is a free editor in the browser; editing costs no credits.
Animate the fuzzy dream dance from first to last frame
On Video, set your standing photo as the first frame and the squat pose as the last, then write the bounce, the arm winds and the knee shake in between.
First frame: standing
Your own photo, unchanged. Animate it uses it as the first frame, so the clip starts exactly as you look.
Last frame: the squat
The pose you designed. The clip has to land on it, so the dance ends low with both hands on your head.
The moves between
Animate it sends both frames to Image to Video, 10 seconds, 9:16, with the moves written in order: a standing bounce, the drop into the squat, a low bounce, arm winds, hands up on the head and a fast knee shake. Pick Seedance 2.5 or Seedance 2.0 Mini, which take a first and a last frame. The frames below come from Stand to squat at 0.4, 2.2, 3.2, 5.0 and 7.8 seconds; that original was animated from its first frame only, so it ends standing with an arm up rather than in the squat.
0.4 sWake up
She starts with her hands over her eyes, mid-yawn, in the sunny bedroom.
2.2 sDeep squat
Down into the wide squat, knees out, fists at the chest.
3.2 sArm winds
Still low, she winds her forearms at chest height on the beat.
5.0 sHands on head
Both hands go up to her head while she stays in the squat.
7.8 sArm up
She stands and throws one arm straight up to finish.
AI clips come back with their own sound. Pick the “I just woke up from a fuzzy dream” sound in TikTok or Reels when you post.
Filming it yourself looks most like you; the AI version takes one photo and a minute, and you never have to hold the squat.
Film it yourself
One photo on Lovart ME
Looks like you
Exactly: your body, your timing
Your face and clothes from the photo; the moves are generated
Time
Take after take until the squat looks right
One photo, a designed pose, then about a minute per clip
Need to dance
Yes: the squat, the arm winds and the knee shake
No: the moves are written in the prompt
Sound
Pick the trend sound while filming
Original music; add the trend sound when you post
When to film it
When you can already do the skeur bounce and want your own timing. People also use the phone-editor (CapCut) template behind the first AI versions, or motion control that copies a dancer's clip onto a photo, which runs 9 to 17 seconds on peer sites.
When to use a photo
When you want the clip today without learning the squat, or want a friend in it who will never film it. A two-photo duo template on Seedance 2.5 runs about 9 seconds elsewhere; here you design the finish pose first, so you choose exactly where the dance lands.
It fixes where the dance lands: the deep squat with both hands on your head. The bounce, the arm winds and the knee shake then have a target, and the clip ends on a pose you already approved instead of wherever the model stops.
What caption do people use?
The line itself: “I just woke up from a fuzzy dream”. It is the opening line of Madonna's “Miles Away” and the words of the trend sound, so one line on its own is enough.
Which photo works?
One adult standing, head to toes in view, because the dance lives in the legs. Wide trousers or shorts move best; good light, nothing over the face. Use your own photo or one of an adult who said yes.
What else is it called?
Fuzzy dream trend, skeur dance, skeur hom, or the I just woke up dance. Skeur is the loose, bouncy groove the moves come from; Kavel traces it to Cape Town's amapiano scene.
Is AI better than filming it?
Filming looks most like you and has your own timing. The AI version takes one photo and about a minute, and you never have to hold the squat or learn the knee shake.
Lovart ME is an independent site and is not affiliated with Madonna, her label, the remix's creator, TikTok, CapCut, MakeViral or Kavel. This page copies only the dance moves: no original recording or dancers' footage. Sample people are original generated adults; the music is original.
