No Scrubs AI 视频生成器
你和两位朋友,一起出演银色布景音乐短片。
了解这个工具
用三张照片制作 No Scrubs AI 好友音乐视频,分别出演中间、左侧与右侧角色;也可选 Solo 只替换中间主角。Wan 3 编辑保留电梯开场、俯拍镜头与三人舞。
2. 上传舞者照片
需要三张单人照。照片 1 对应最终正面舞蹈的中间主角,照片 2 对应左侧,照片 3 对应右侧。跨镜头始终绑定原角色,不按特写出场顺序分配。
照片 1 · 中间 / 电梯主角必填
照片 2 · 左侧角色必填
照片 3 · 右侧角色必填
每张照片一人,面部清晰,尽量展示想穿的服装;提示词要求使用照片里的头发与穿搭。JPG、PNG 或 WebP,每张最多 10 MiB,至少 300 × 300 像素。
本模板共 600 积分,包含全部所需照片。失败或超时退还实际扣费。
预览 No Scrubs AI 场景
这里展示换人前的导入参考片段。单人版只替换中间主角,另外两位保留原片外观与各自的个人特写。
- 01选择 Trio、Solo 或 Guys,先预览原始场景。
- 02三人版依次上传中间、左侧、右侧照片,单人版上传一张。
- 03分享前检查跨镜头身份、穿搭、舞步与音轨。
选择 No Scrubs AI 出演阵容
参考片段导入自 MakeViral(makeviral.ai),并非 ImageLayered 实测输出。Trio 与 Guys 分别使用参考站对应的 AI 示例作为编辑输入。单人版复用 Trio 原片,生成 15 秒主角编辑,预览仍显示原始阵容。参考站将配乐描述为原创音轨,并非 TLC 原曲。
Trio · 好友三人版
三张照片,分别出演银色布景中央、左侧与右侧。
3 张照片 · 15 秒
来源:MakeViral · 导入参考视频Solo · 单人主角
一张照片只替换中间主角,另两位及各自特写保留原样。
1 张照片 · 15 秒
来源:MakeViral · 导入参考视频Guys · 男生参考场景
三张照片,编辑三位男士的参考场景。
3 张照片 · 15 秒
来源:MakeViral · 导入参考视频每个版本都有内置 Wan 3 编辑提示词
无需填写提示词。Trio 与 Guys 在六个镜头中分别绑定三张照片。Solo 只替换原片中间主角:电梯开场、第二段个人特写、俯拍中央与群舞中间人物。另外两位角色及各自的个人特写保持原样。提示词分别约束面部、头发、穿搭与配饰;AI 仍可能改变面部或动作。
Trio · 好友三人版
Image 1 is CHARACTER_MIDDLE, replacing the dark-haired woman in a black camisole and loose blue jeans: alone in the elevator, in the middle performer's own close-up, in the center of the overhead shot and MIDDLE in the final frontal dance. Image 2 is CHARACTER_LEFT, replacing the curly-haired woman in a gray long-sleeve top and blue jeans: the first pointing close-up and LEFT in the final frontal dance. Image 3 is CHARACTER_RIGHT, replacing the blond man in a navy T-shirt and blue jeans: the finger-wag close-up and RIGHT in the final frontal dance. Permanently bind each Image to that original person in every shot. An individual close-up must use its own character's Image, never another Image. Never blend identities or swap the three assignments. SHOT ASSIGNMENTS IN THE ACTUAL TRIO SOURCE: 1. 0.00-2.75s, silver elevator: CHARACTER_MIDDLE alone -> Image 1. 2. 2.75-4.875s, curly-haired woman in gray pointing toward the lens: CHARACTER_LEFT alone -> Image 2, NOT Image 1. 3. 4.875-6.708s, dark-haired woman in a black camisole raising both hands: CHARACTER_MIDDLE alone -> Image 1. 4. 6.708-8.667s, blond man in navy leaning in and wagging a finger: CHARACTER_RIGHT alone -> Image 3, NOT Image 1. 5. 8.667-10.542s, overhead fisheye: center CHARACTER_MIDDLE -> Image 1; original gray-top CHARACTER_LEFT -> Image 2; original navy-shirt CHARACTER_RIGHT -> Image 3. Keep all three identities distinct even when the lens distorts faces. 6. 10.542-15.00s, group dance: final frontal order is Image 2 LEFT, Image 1 MIDDLE, Image 3 RIGHT. Follow each original dancer through the turn and preserve each dancer's own steps and gestures; do not reassign by temporary screen position. For each assigned character, use ONLY that character's assigned Image for identity: face shape, eyes and eye color, brows, nose, lips, skin tone, apparent age, hair color, hairline, hairstyle and length, facial hair, glasses, jewelry and other visible accessories. Keep that identity recognizable and consistent in wide shots, close-ups, profiles, turns and partial occlusions. Fully replace the assigned source person's appearance, hair, body proportions and clothing with the reference person's own appearance. Fit the same reference outfit naturally across every shot, without borrowing another performer's face, hair or clothes. If lower-body clothes are outside the photo, extend the visible outfit coherently with simple neutral trousers and shoes; do not retain distinctive source clothes. Do not import photo backgrounds. Preserve expression and lip-motion timing using the replacement person's own facial anatomy. No blended faces, source-face residue, duplicate identities, extra people, translucent bodies or extra limbs. Edit Video 1 as a selective character replacement. Keep the existing footage's first fifteen seconds, original cuts and shot lengths, gestures, choreography, gaze, mouth movement, expressions, camera movement, lens distortion, silver set, cool lighting, shadow placement, reflection geometry, framing and source audio. Match each replaced character’s floor reflections to that assigned person and outfit, retaining their original motion and lighting; keep protected people’s reflections unchanged. Do not recreate the scene or generate a new soundtrack. Keep exactly the source number of visible people in each shot: one in each elevator or close-up shot, three in each overhead or group-dance shot. Do not insert side dancers into individual shots, remove them from group shots or add text, captions, logos or watermarks. Track the same original person across cuts and turns, never by changing screen position. The shot times below locate existing cuts approximately; preserve the actual cuts and use the original person's identity as the authority at shot boundaries.
Solo · 单人主角
Image 1 defines the ONLY lead identity to insert. REPLACE ONLY SOURCE_MIDDLE: the dark-haired woman with loose dark hair, a black camisole and loose blue jeans. She is alone in the opening elevator, appears in the SECOND individual close-up raising both hands, is centered in the overhead shot and is MIDDLE in the final frontal group dance. Replace this one original person wherever she appears, using Image 1 for her entire appearance and outfit. PROTECTED SOURCE CHARACTERS, WITH NO ASSIGNED IMAGE: SOURCE_LEFT is the curly-haired woman in a gray long-sleeve top and blue jeans. Preserve her original face, hair, skin, body, gray outfit and identity in EVERY appearance, including her FIRST pointing close-up, overhead appearance and dance. SOURCE_RIGHT is the blond man in a navy T-shirt and blue jeans. Preserve his original face, hair, skin, body, navy outfit and identity in EVERY appearance, including his THIRD finger-wag close-up, overhead appearance and dance. Never apply Image 1 to SOURCE_LEFT or SOURCE_RIGHT, including when either fills the whole frame or occupies the screen center. A close-up is a camera framing, not permission to replace a different character. Never clone the lead's face, hair, clothes or accessories onto either protected person. The two protected people remain distinct, original source characters, with their own featured close-ups; they are not restricted to backup-dancer appearances in group shots. SHOT ASSIGNMENTS IN THE ACTUAL SOLO INPUT (THE TRIO SOURCE): 1. 0.00-2.75s, silver elevator: replace ONLY SOURCE_MIDDLE with Image 1. 2. 2.75-4.875s, SOURCE_LEFT's pointing close-up: KEEP THIS ENTIRE SHOT UNEDITED. This is the gray-top woman's own face and clothes, not the uploaded person. 3. 4.875-6.708s, SOURCE_MIDDLE's raised-hands close-up: replace ONLY SOURCE_MIDDLE with Image 1. 4. 6.708-8.667s, SOURCE_RIGHT's finger-wag close-up: KEEP THIS ENTIRE SHOT UNEDITED. This is the blond man's own face and clothes, not the uploaded person. 5. 8.667-10.542s, overhead fisheye: replace ONLY center SOURCE_MIDDLE with Image 1; keep SOURCE_LEFT and SOURCE_RIGHT original and distinct, including their visible faces, hair, outfits and gestures. 6. 10.542-15.00s, group dance: replace ONLY original middle dancer SOURCE_MIDDLE with Image 1. Keep the original gray-top woman and navy-shirt man unchanged through the turn and ending. Follow original identity rather than screen position. Exactly ONE uploaded identity appears in a group shot. APPEARANCE INSTRUCTIONS APPLY ONLY TO SOURCE_MIDDLE: For each assigned character, use ONLY that character's assigned Image for identity: face shape, eyes and eye color, brows, nose, lips, skin tone, apparent age, hair color, hairline, hairstyle and length, facial hair, glasses, jewelry and other visible accessories. Keep that identity recognizable and consistent in wide shots, close-ups, profiles, turns and partial occlusions. Fully replace the assigned source person's appearance, hair, body proportions and clothing with the reference person's own appearance. Fit the same reference outfit naturally across every shot, without borrowing another performer's face, hair or clothes. If lower-body clothes are outside the photo, extend the visible outfit coherently with simple neutral trousers and shoes; do not retain distinctive source clothes. Do not import photo backgrounds. Preserve expression and lip-motion timing using the replacement person's own facial anatomy. No blended faces, source-face residue, duplicate identities, extra people, translucent bodies or extra limbs. Edit Video 1 as a selective character replacement. Keep the existing footage's first fifteen seconds, original cuts and shot lengths, gestures, choreography, gaze, mouth movement, expressions, camera movement, lens distortion, silver set, cool lighting, shadow placement, reflection geometry, framing and source audio. Match each replaced character’s floor reflections to that assigned person and outfit, retaining their original motion and lighting; keep protected people’s reflections unchanged. Do not recreate the scene or generate a new soundtrack. Keep exactly the source number of visible people in each shot: one in each elevator or close-up shot, three in each overhead or group-dance shot. Do not insert side dancers into individual shots, remove them from group shots or add text, captions, logos or watermarks. Track the same original person across cuts and turns, never by changing screen position. The shot times below locate existing cuts approximately; preserve the actual cuts and use the original person's identity as the authority at shot boundaries. Final scope check: only SOURCE_MIDDLE receives Image 1. Keep the other two characters and both of their individual close-up shots unchanged.
Guys · 男生参考场景
Image 1 is CHARACTER_MIDDLE, replacing the blond man: alone in the silver elevator, in the second individual close-up, center in the overhead shot and MIDDLE in the final frontal dance. In the source he wears a navy T-shirt in individual shots and a dark track jacket with light sleeve stripes in the final dance; those are the SAME person and must use Image 1's same appearance and outfit. Image 2 is CHARACTER_LEFT, replacing the curly-haired bearded man in navy: the first pointing close-up and LEFT in the final frontal dance. Image 3 is CHARACTER_RIGHT, replacing the close-cropped bearded man in a red-and-black plaid overshirt: the third close-up and RIGHT in the final frontal dance. Do not assign photos by close-up order. Never confuse the blond man with the curly-haired bearded man merely because both wear navy tops. Use each uploaded person's own appearance and clothes regardless of gender, keeping three separate identities. Source costume changes must not change the mapping. SHOT ASSIGNMENTS IN THE ACTUAL GUYS SOURCE: 1. 0.00-2.50s, elevator: blond CHARACTER_MIDDLE alone -> Image 1. 2. 2.50-5.10s, curly-haired bearded pointing close-up: CHARACTER_LEFT alone -> Image 2, NOT Image 1. 3. 5.10-7.533s, blond raised-hands close-up: CHARACTER_MIDDLE alone -> Image 1. 4. 7.533-9.467s, close-cropped bearded man in plaid, finger-wag close-up: CHARACTER_RIGHT alone -> Image 3, NOT Image 1. 5. 9.467-11.467s, overhead: center blond CHARACTER_MIDDLE -> Image 1; curly-haired bearded CHARACTER_LEFT -> Image 2; plaid CHARACTER_RIGHT -> Image 3. 6. 11.467-15.00s, group dance: frontal order is Image 2 LEFT, Image 1 MIDDLE, Image 3 RIGHT. Image 1 replaces the blond man now wearing the striped-sleeve jacket. Follow each original person's own motion through turns without exchanging identities. For each assigned character, use ONLY that character's assigned Image for identity: face shape, eyes and eye color, brows, nose, lips, skin tone, apparent age, hair color, hairline, hairstyle and length, facial hair, glasses, jewelry and other visible accessories. Keep that identity recognizable and consistent in wide shots, close-ups, profiles, turns and partial occlusions. Fully replace the assigned source person's appearance, hair, body proportions and clothing with the reference person's own appearance. Fit the same reference outfit naturally across every shot, without borrowing another performer's face, hair or clothes. If lower-body clothes are outside the photo, extend the visible outfit coherently with simple neutral trousers and shoes; do not retain distinctive source clothes. Do not import photo backgrounds. Preserve expression and lip-motion timing using the replacement person's own facial anatomy. No blended faces, source-face residue, duplicate identities, extra people, translucent bodies or extra limbs. Edit Video 1 as a selective character replacement. Keep the existing footage's first fifteen seconds, original cuts and shot lengths, gestures, choreography, gaze, mouth movement, expressions, camera movement, lens distortion, silver set, cool lighting, shadow placement, reflection geometry, framing and source audio. Match each replaced character’s floor reflections to that assigned person and outfit, retaining their original motion and lighting; keep protected people’s reflections unchanged. Do not recreate the scene or generate a new soundtrack. Keep exactly the source number of visible people in each shot: one in each elevator or close-up shot, three in each overhead or group-dance shot. Do not insert side dancers into individual shots, remove them from group shots or add text, captions, logos or watermarks. Track the same original person across cuts and turns, never by changing screen position. The shot times below locate existing cuts approximately; preserve the actual cuts and use the original person's identity as the authority at shot boundaries.
如何与朋友制作 No Scrubs AI 视频
选择版本、分配角色,一起出演银色布景音乐短片。
1. 选择 Trio、Solo 或 Guys
先看参考场景,Trio 与 Guys 需三张照片,Solo 需一张。所有版本均输出 15 秒。
2. 分配照片并生成
三人版依次上传中间、左侧、右侧角色照片。选择 480p、720p 或 1080p,确认报价后用 Wan 3 生成。
3. 检查并下载
等待时可玩 2048。完成后检查跨镜头人物对应、服装、舞步与声音,再下载 MP4,成片保留 30 天。
No Scrubs AI 常见问题
关于照片顺序、单人和三人版本、Wan 3、积分与下载。
什么是 No Scrubs AI?
这类照片生成视频以 90 年代末 R&B 音乐视频为灵感,包含银色电梯、个人特写、俯拍与三人舞。本工具使用 Wan 3,将你的照片编辑进导入的 AI 参考片段。
如何和朋友一起制作 No Scrubs AI 视频?
选择 Trio 或 Guys,上传三张单人照。照片 1 对应最终正面舞蹈的中间角色并乘电梯开场,照片 2 对应左侧,照片 3 对应右侧。映射跟随原人物跨镜头保持,不按特写出场顺序分配。
可以一个人做,或者选择男生版吗?
可以。Solo 只需一张照片,仅替换原片中间主角,覆盖电梯开场、主角自己的特写和群像中间角色。另两位原片人物保留自己的面部、服装与个人特写。Guys 使用三位男士的参考片段,需要三张照片。三人版本均要求使用上传照片中的外观,也支持混合阵容。
时长、比例与分辨率是什么?
三个版本均为 15 秒、9:16 竖屏,包括本站单人版。支持 480p、720p、1080p,下载为 MP4。本站未采用参考站 10 秒单人版的生成流程。
需要多少积分?
三个版本价格相同:480p 为 600 积分,720p 为 1200,1080p 为 2400。包含所有所需照片及完整 15 秒成片。失败或超时退还实际扣费。
会保留我的服装吗?有 TLC 原曲吗?
提示词要求使用每人照片中的面部、头发与服装,配置保留原片音轨,不生成新音轨。MakeViral 将导入示例的配乐描述为原创音轨,并非官方 TLC No Scrubs 原曲。AI 仍可能改变面部、服装与舞步,请检查完整结果。
能恢复任务、稍后下载吗?
上传与免费排队期间请保持页面打开。已受理任务会在后台继续,可在 24 小时内恢复最新任务,成片在最近生成中保留 30 天。等待时可玩本地 2048。
