动漫角色讲解 AI 视频学校概念广告

红发角色、青粉图形与日文标题组成虚构学校的宣传视频。

生成结果

来源未标注生成方式

输入参考素材

0 / 1 个文件已公开

按提示词中的编号使用素材；输入参考与上方生成结果分开展示。

来源未公开全部参考文件；复现时请为缺失输入准备本人或已获授权的素材。

Picture 1 · 来源未提供文件

复现时，仅使用本人或已获授权的肖像、声音及其他素材。

提示词信息
分享者	@tanabe_fragm
模型	MiniMax H3
分类	广告与电商
数据集	SkyNotSilent · awesome-MiniMax-H3-cases

完整提示词

subject_definitions:
<Subject 1> is the anime girl in <Picture 1>: Japanese TV-anime style, late teens, long vivid
red hair with loose flowing strands, large emerald-green eyes, white hoodie over a white top.
<Subject 2> is the graphic style in <Picture 1>: crisp flat 2D vector overlays on white -
cyan-turquoise and magenta-pink triangles, dot grids, diagonal speed shards, glowing UI
panels, thin rounded icons.
<Subject 3> is the school wordmark in <Picture 1>: the Japanese logotype "デジラク" in a heavy
rounded gothic face, teal-green with a soft drop shadow.

summary:
[reference generation] A 15-second Japanese commercial for an online AI-video creator school,
drawn as clean anime-style 2D animation over fast smooth motion graphics with animated
Japanese type, carried by a female voiceover. <Subject 1> discovers AI video, learns, and
invites the viewer to start. <Picture 1> is a storyboard contact sheet used only for casting
and style.

retention_analysis:
<Subject 1> (appears in [Shot 1], [Shot 3], [Shot 5], [Shot 8] and [Shot 10]):
partially_preserved - her red hair, emerald-green eyes, face and white hoodie stay identical;
only pose and expression change.
<Subject 2> (appears in every shot): weak_reference - only the white palette, flat vector feel
and speed-shard energy carry over; the storyboard's grid, borders and numbers are not
reproduced.
<Subject 3> (appears in [Shot 9]): fully_preserved - the letterforms, teal-green color and
proportions are kept exactly.

detailed_description:
[Shot 1] Clean Japanese TV-anime 2D animation over crisp flat vector motion graphics with
animated Japanese type, cut like a fast bright commercial. On a white field <Subject 1> runs
in from the right toward camera, hair streaming, as cyan and magenta shards burst past her.
Charcoal type slides in line by line on the left reading "AIで、アイデアが動き出す！",
"動き出す" snapping to double size in cyan. A bright Japanese woman in her early twenties with
a light buoyant friendly voice (S1) says off-screen:
<d>[Japanese] AIで、アイデアが動き出す。</d> Push In, small amplitude, fast speed.
[Shot 2] At 00:01.500, a cyan shard wipes across. A dozen glowing video panels fly through
deep blue space in one curved stream, each holding a different anime, cityscape or fantasy
scene. White type assembles in two stages at the lower left reading
"こんな動画も…". (S1): <d>[Japanese] こんな動画も。</d> Truck Right, large amplitude, fast
speed.
[Shot 3] At 00:03.000, a white field wipes in. Close on <Subject 1>, hands clasped at her
chest, eyes widening into a sparkle as four-point stars pop in around her. Bold type
scales up hard on the left reading "私にも 作れるかも！", "私" landing in magenta-pink.
<Subject 1>, a younger softer close-mic voice (S2), says quietly to herself:
<d>[Japanese] 私にも、作れるかも。</d> Push In, small amplitude, slow speed.
[Shot 4] At 00:04.500, a clean white field wipes in. Charcoal type draws itself across the top
reading "デジラクで学べること". Three thin rounded icons then pop in one at a time across the
middle - a laptop, an AI chip, a lightbulb - each with a label rising
beneath: "AI動画の基礎", "最新AIツール", "アイデアの発想法". No voice here. Static Shot.
[Shot 5] At 00:06.000, a cyan shard wipes in. <Subject 1> sits at a wide monitor, typing, as
glowing UI panels, timelines and waveform bars snap into place around the screen. A teal
banner sweeps in from the left with white type reading "ゼロからでも大丈夫！". (S1):
<d>[Japanese] ゼロからでも、大丈夫。</d> Truck Left, small amplitude, slow speed.
[Shot 6] At 00:07.500, diagonal cyan and magenta bands sweep across a white field. Huge
charcoal type slams in from both sides and collides at center reading
"プロの現場の技術が身につく！", "プロの現場" in cyan with a thick underline drawing itself as
light shards scatter. (S1): <d>[Japanese] プロの技術が、身につく。</d> Static Shot.
[Shot 7] At 00:09.000, the frame splits along one steep diagonal: a dull gray unfinished panel
on the left slides toward a bright finished anime artwork on the right as a glowing arrow
pushes between them. White type streaks in along the diagonal in two stages reading
"あなたの作品が、" then "ここから変わる。". (S1): <d>[Japanese] あなたの作品が、変わる。</d>
Push In, small amplitude, fast speed.
[Shot 8] At 00:10.500, a bright sky-blue field with radiating light wipes in. <Subject 1>
reaches one hand straight toward camera, smiling wide, as cyan and magenta shards streak past
her. Bold white type assembles in two stages on the right reading "一緒に、" then "つくろう！".
(S2), now bright and open: <d>[Japanese] 一緒に、つくろう！</d> Pull Out, small amplitude,
fast speed.
[Shot 9] At 00:12.000, a clean white field wipes in. Small charcoal type fades up across the
top reading "AI動画クリエイター育成スクール". Scattered letters then fly in and lock together
into <Subject 3> at center, a light sweep passing over it. Three thin rounded
icons pop in one at a time across the lower third - a document, a person, a group - with
labels beneath: "実践課題", "プロが指導", "コミュニティ". (S1):
<d>[Japanese] AI動画スクール、デジラク。</d> Static Shot.
[Shot 10] At 00:13.500, a white field with cyan and magenta shards wipes in. <Subject 1>
stands on the right pointing forward toward camera with a bright smile. Bold charcoal type
scales up on the left reading "今すぐ無料体験へ！", and a rounded teal button assembles beneath
reading "公式サイトはこちら", pulsing once. (S1), a half step warmer and more forward:
<d>[Japanese] 今すぐ、無料体験へ。</d> The frame holds still until exactly 15.000 seconds.
Static Shot.
Throughout: keep one clean Japanese TV-anime style with crisp linework, soft cel shading and
bright color - never photorealistic, never 3D-rendered. Only (S1) and (S2) speak, exactly the
lines above and nothing else: no ad-libs, no extra narration, no singing, no lip sync except
in [Shot 3] and [Shot 8]. Every on-screen line is bold rounded Japanese gothic type that
builds as animated shapes - never a movie subtitle, caption strip, letterboxed caption, footer
text or romaji. No storyboard grid, borders, numbers or contact-sheet layout, and no split
screen except the diagonal in [Shot 7]. Keep <Subject 1>'s red hair and emerald-green eyes
identical, cut sharp with no dissolves, keep motion smooth, and hold the palette to
crisp white, cyan-turquoise, magenta-pink, sky blue and charcoal.

overall_soundscape:
Bright airy whooshes carry the shards and flying panels, and soft digital clicks mark each
icon and UI element as it snaps into place. A rising sparkle lifts as her eyes widen, a solid
impact lands when the type collides, and keyboard taps run under the editing shot. A
clean chime rings once on the button at the end.

non_diegetic_music:
An original 15-second bright uplifting Japanese commercial track at a fast tempo: plucked
synth lead, four-on-the-floor kick, tight claps, airy pads, bells, mixed low under the voiceover. It opens at full drive, drops the drums for one bar at 00:03.000 leaving
only a sparkling bell, snaps back at 00:04.500 and builds, lifts into a rising sweep at
00:10.500, opens widest at 00:12.000, then lands on one bright sustained chord that decays
cleanly at 15.000 seconds.

在 DraPub 中尝试

看看同一视觉方向里的其他提示词。

返回提示词画廊