← 返回 Skills 市场
peand-rover

Ai Video Background

作者 peandrover adam · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ pending
113
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install ai-video-background
功能描述
Change, replace, and customize video backgrounds with AI — swap backgrounds in real-time, add virtual environments, blur distracting backgrounds, replace gre...
使用说明 (SKILL.md)

AI Video Background — Record Anywhere. Look Like You Are Anywhere.

The background of a video communicates as much as the speaker. A messy bedroom behind a business presenter undermines credibility. A plain white wall behind a creative professional feels sterile. A busy café behind a course instructor distracts from the lesson. The background sets the context, establishes professionalism, and either supports or sabotages the content. Traditional solutions — dedicated studios, green screen setups, careful location selection — add cost, time, and logistical constraints to every recording. A basic green screen setup costs $100-500 and requires proper lighting to avoid spill. A dedicated studio space costs $500-5,000/month. Even virtual backgrounds in Zoom require strong hardware and often produce visible artifacts around hair and edges. NemoVideo changes backgrounds after recording. Film anywhere — bedroom, kitchen, hotel room, car, café — and replace the background in post. The AI segments the foreground subject (people, objects) from the background with edge precision that handles hair, translucent elements, and moving subjects. Replace with: a professional office, a custom branded backdrop, a cinematic environment, a blurred version of the original (focus on the speaker), or any image or video background.

Use Cases

  1. Remote Worker — Professional Backdrop for Video Calls (any length) — A consultant records client-facing videos from their apartment. The background shows a bed, laundry, and a cat. NemoVideo: segments the consultant from the background with precise edge detection (including hair strands), replaces with a clean modern office background (bookshelf, plant, soft lighting), color-matches the subject to the new environment (lighting consistency), and maintains natural-looking edges throughout the video. The consultant looks like they have a professional studio. The cat remains out of frame.

  2. Content Creator — Themed Backgrounds for Series (multiple) — A creator produces different content series that each need a distinct visual identity: tech reviews (dark studio with RGB lighting), cooking tips (kitchen environment), motivation (sunrise mountain backdrop). NemoVideo: takes all recordings from the same desk setup and applies different backgrounds per series, maintaining consistent lighting integration for each environment. One recording space, multiple visual identities.

  3. Corporate Training — Branded Backdrop (batch) — A company produces 20 training videos with 8 different presenters. Each presenter records in their own location (home offices, meeting rooms, hotel rooms). NemoVideo: applies the company's branded background to all 20 videos (consistent backdrop with company logo, brand colors, subtle animated elements), color-matches each presenter to the branded environment, and produces a visually consistent training library. Eight different locations become one professional brand.

  4. Background Blur — Focus on the Speaker (any length) — A presenter records in a busy office. Colleagues walking behind, screens with sensitive information visible, visual clutter competing with the speaker. NemoVideo: detects the speaker as the foreground subject, applies cinematic background blur (adjustable bokeh intensity), keeps the speaker in sharp focus, and maintains natural depth-of-field appearance. The background becomes a soft, non-distracting wash of color while the speaker commands full attention.

  5. Green Screen Replacement — Without the Green Screen (any length) — A creator wants to place themselves in front of a fantasy landscape for a storytelling video but does not own a green screen. NemoVideo: segments the creator from their actual background using AI (no green screen needed), replaces with the fantasy landscape image or video, handles edge blending around hair and clothing, adds subtle environmental effects (matching the landscape's lighting on the subject's skin), and exports. The green screen effect without the green screen — from any recording environment.

How It Works

Step 1 — Upload Video

Any video with a subject you want to separate from the background. Works with: single person, multiple people, objects, or any defined foreground element.

Step 2 — Choose Background Treatment

Replace (new image/video background), blur (adjustable intensity), remove (transparent for compositing), or brand (company backdrop).

Step 3 — Generate

curl -X POST https://mega-api-prod.nemovideo.ai/api/v1/generate \
  -H "Authorization: Bearer $NEMO_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "skill": "ai-video-background",
    "prompt": "Replace the background of a 5-minute talking-head video. Current background: messy home office. New background: clean modern office with bookshelf, plant, and soft window light from the left. Match the new background lighting to the subject (light source from upper-left). Maintain precise hair edge detection. Also generate a blurred-background version (cinematic bokeh, moderate intensity) as an alternative. Export both at 16:9 1080p.",
    "background_action": "replace",
    "new_background": "modern-office-bookshelf-window-light",
    "lighting_match": "upper-left",
    "edge_quality": "hair-precise",
    "alternatives": [
      {"action": "blur", "intensity": "moderate-cinematic-bokeh"}
    ],
    "format": "16:9",
    "resolution": "1080p"
  }'

Step 4 — Review Edge Quality

Watch for: hair edge artifacts, shadow consistency, lighting match. Adjust if needed. Download preferred version.

Parameters

Parameter Type Required Description
prompt string Background change description
background_action string "replace", "blur", "remove", "brand"
new_background string Background description, image URL, or preset
blur_intensity string "light", "moderate", "heavy", "cinematic-bokeh"
lighting_match string Match light direction to new background
edge_quality string "standard", "hair-precise", "maximum"
brand_background object {logo, colors, elements} for corporate backdrop
alternatives array Generate multiple background versions
format string "16:9", "9:16", "1:1"
resolution string "720p", "1080p", "4K"
batch array Multiple videos with same background

Output Example

{
  "job_id": "avbg-20260328-001",
  "status": "completed",
  "source_duration": "5:08",
  "segmentation": {
    "subject": "single person",
    "edge_quality": "hair-precise",
    "frames_processed": 9144
  },
  "outputs": {
    "replaced": {
      "file": "video-office-bg.mp4",
      "background": "modern office (bookshelf + window light)",
      "lighting_matched": true
    },
    "blurred": {
      "file": "video-blur-bg.mp4",
      "blur": "moderate cinematic bokeh"
    }
  }
}

Tips

  1. Lighting direction must match between subject and new background — If the subject is lit from the left but the new background has a window on the right, the scene looks wrong even if the viewer cannot articulate why. Always match light direction.
  2. Hair edge quality is the test of professional vs. amateur background replacement — Blocky edges around hair instantly reveal the replacement. Hair-precise edge detection handles individual strands and creates the natural blending that makes the background change invisible.
  3. Background blur is the safest professional option — If the goal is simply "look professional," blurring the existing background is often better than replacing it. It looks natural (because it IS the real environment, just defocused), requires no lighting matching, and signals "intentional focus on the speaker" rather than "virtual background."
  4. Branded backgrounds create instant corporate consistency — When every presenter in a 20-person team uses the same branded background, the video library looks cohesive regardless of where each person recorded. Brand consistency from distributed recording locations.
  5. Test with fast movement — The edge segmentation is most visible when the subject moves quickly (gesturing, turning). Always preview a section with movement to verify edge quality before processing the full video.

Output Formats

Format Resolution Use Case
MP4 16:9 1080p / 4K YouTube / website / presentation
MP4 9:16 1080x1920 TikTok / Reels
MP4 1:1 1080x1080 Instagram / LinkedIn
MOV (alpha) 1080p Compositing (transparent BG)

Related Skills

安全使用建议
This skill appears internally consistent: it uploads videos to a NemoVideo API using a NEMO_TOKEN and replaces/blur backgrounds as described. Before installing or enabling it, consider: 1) Privacy: recorded videos (and anything visible in frames) will be sent to a third-party endpoint — confirm the service's data retention, processing, and deletion policies. 2) Credentials: provide a service-specific NEMO_TOKEN (do not reuse high-privilege tokens such as cloud provider or GitHub tokens). 3) Metadata mismatch: the registry summary says no required env vars while the skill uses NEMO_TOKEN and a config path (~/.config/nemovideo/); ask the publisher how the token is expected to be supplied (env var vs config file) and whether the skill will read that config path automatically. 4) Test safely: try with non-sensitive sample videos first. 5) Network trust: verify the endpoint (nemovideo.ai) and confirm you trust the operator before sending sensitive content.
能力评估
Purpose & Capability
The skill is an instruction-only wrapper for a video-background API (NemoVideo). Requesting a NEMO_TOKEN and performing an HTTPS POST to a nemovideo.ai endpoint is consistent with the described capability.
Instruction Scope
SKILL.md contains a curl example that uploads videos to https://mega-api-prod.nemovideo.ai/api/v1/generate with an Authorization: Bearer $NEMO_TOKEN. The instructions do not ask the agent to read unrelated local files, but they do direct uploading potentially sensitive video content to an external service — this is expected but privacy-sensitive.
Install Mechanism
No install spec or code is provided (instruction-only). That reduces on-disk risk — the main runtime risk is network transmission of user video data to the external API.
Credentials
The declared primary credential (NEMO_TOKEN) is appropriate for calling the NemoVideo API. However, the registry summary listed 'Required env vars: none' while primaryEnv is NEMO_TOKEN, and metadata lists a config path (~/.config/nemovideo/). This small inconsistency should be clarified (the skill will need the token — either via env var or config file).
Persistence & Privilege
The skill is not marked 'always: true' and has no install steps that modify other skills or system-wide settings. The agent may call it autonomously (default), which is normal; combine that with credential use consciously.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install ai-video-background
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /ai-video-background 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release — AI-powered video background changer for post-production. - Change, replace, and customize video backgrounds with AI segmentation. - Swap backgrounds in real-time, blur distractions, or insert custom images/videos — no green screen required. - Handles edge precision (hair, objects) and lighting adjustments for natural results. - Supports multiple use cases: professional, creative, branded, and privacy-focused backgrounds. - API allows batch processing, background alternatives, and adjustable output formats/resolution.
元数据
Slug ai-video-background
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Ai Video Background 是什么?

Change, replace, and customize video backgrounds with AI — swap backgrounds in real-time, add virtual environments, blur distracting backgrounds, replace gre... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 113 次。

如何安装 Ai Video Background?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install ai-video-background」即可一键安装,无需额外配置。

Ai Video Background 是免费的吗?

是的,Ai Video Background 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Ai Video Background 支持哪些平台?

Ai Video Background 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Ai Video Background?

由 peandrover adam(@peand-rover)开发并维护,当前版本 v1.0.0。

💬 留言讨论