modelslab-interior-design

ModelsLab インテリアデザイン

AI を活用したインテリアデザイン、部屋の装飾、間取り図の作成、建物外観の修復。

このスキルを使用する場面

• インテリア空間を再設計する
• AI の支援で部屋を装飾する
• 画像から間取り図を作成する
• 部屋のスタイルや雰囲気を変える
• 建物の外観を修復または改善する
• デザインのモックアップやバリエーションを作成する
• リノベーションのアイデアを視覚化する

利用可能なエンドポイント

インテリアデザイン

POST https://modelslab.com/api/v6/interior/interior

ルームデコレーター

POST https://modelslab.com/api/v6/interior/room_decorator

間取り図の作成

POST https://modelslab.com/api/v6/interior/floor_planning

外観の修復

POST https://modelslab.com/api/v6/interior/exterior_restorer

シナリオ変更

POST https://modelslab.com/api/v6/interior/scenario_changer

オブジェクトの削除

POST https://modelslab.com/api/v6/interior/object_removal

インテリアミキサー

POST https://modelslab.com/api/v6/interior/interior_mixer

インテリアの再設計

import requests

def redesign_interior(room_image, design_prompt, api_key):
    """Redesign an interior space based on a prompt.

    Args:
        room_image: URL of the room photo
        design_prompt: Description of desired design
        api_key: Your ModelsLab API key

    Returns:
        URL of the redesigned interior
    """
    response = requests.post(
        "https://modelslab.com/api/v6/interior/interior",
        json={
            "key": api_key,
            "init_image": room_image,
            "prompt": design_prompt,
            "negative_prompt": "low quality, distorted, unrealistic",
            "num_inference_steps": 31,  # 21, 31, or 41
            "guidance_scale": 7.5,
            "strength": 0.7
        }
    )

    data = response.json()

    if data["status"] == "success":
        return data["output"][0]
    else:
        raise Exception(f"Error: {data.get('message', 'Unknown error')}")

# Usage
redesigned = redesign_interior(
    "https://example.com/living-room.jpg",
    "Modern minimalist living room with Scandinavian furniture, white walls, natural light",
    "your_api_key"
)
print(f"Redesigned room: {redesigned}")

ルームデコレーター

def decorate_room(room_image, decor_prompt, api_key, specific_object=None):
    """Decorate a room with AI-generated furniture and decor.

    Args:
        room_image: URL of the empty or basic room
        decor_prompt: Description of desired decoration
        specific_object: Specific furniture/decor item that must appear
    """
    payload = {
        "key": api_key,
        "init_image": room_image,
        "prompt": decor_prompt,
        "negative_prompt": "cluttered, low quality, distorted",
        "num_inference_steps": 31,
        "guidance_scale": 7.5,
        "strength": 0.8
    }

    if specific_object:
        payload["specific_object"] = specific_object

    response = requests.post(
        "https://modelslab.com/api/v6/interior/room_decorator",
        json=payload
    )

    data = response.json()

    if data["status"] == "success":
        return data["output"][0]
    else:
        raise Exception(data.get("message"))

# Decorate empty room
decorated = decorate_room(
    "https://example.com/empty-room.jpg",
    "Cozy bedroom with warm lighting, plants, wooden furniture",
    "your_api_key",
    specific_object="king size bed"
)
print(f"Decorated room: {decorated}")

間取り図の作成

def generate_floor_plan(room_image, api_key):
    """Generate a floor plan from a room image.

    Args:
        room_image: URL of room photo
        api_key: Your API key

    Returns:
        URL of the generated floor plan
    """
    response = requests.post(
        "https://modelslab.com/api/v6/interior/floor_planning",
        json={
            "key": api_key,
            "init_image": room_image
        }
    )

    data = response.json()

    if data["status"] == "success":
        return data["output"][0]
    else:
        raise Exception(data.get("message"))

# Generate floor plan
floor_plan = generate_floor_plan(
    "https://example.com/room-photo.jpg",
    "your_api_key"
)
print(f"Floor plan: {floor_plan}")

外観の修復

def restore_exterior(building_image, restoration_prompt, api_key):
    """Restore or enhance building exterior.

    Args:
        building_image: URL of building exterior photo
        restoration_prompt: Description of desired restoration
    """
    response = requests.post(
        "https://modelslab.com/api/v6/interior/exterior_restorer",
        json={
            "key": api_key,
            "init_image": building_image,
            "prompt": restoration_prompt,
            "negative_prompt": "damaged, old, worn",
            "num_inference_steps": 31,
            "guidance_scale": 7.5
        }
    )

    data = response.json()

    if data["status"] == "success":
        return data["output"][0]
    else:
        raise Exception(data.get("message"))

# Restore old building
restored = restore_exterior(
    "https://example.com/old-building.jpg",
    "Restored Victorian house with fresh paint, new windows, landscaped garden",
    "your_api_key"
)

シナリオ変更

def change_room_scenario(room_image, new_scenario, api_key):
    """Change the environment scenario of a room.

    Args:
        room_image: URL of room photo
        new_scenario: Description of new scenario/ambiance
    """
    response = requests.post(
        "https://modelslab.com/api/v6/interior/scenario_changer",
        json={
            "key": api_key,
            "init_image": room_image,
            "prompt": new_scenario,
            "num_inference_steps": 31,
            "guidance_scale": 7.5
        }
    )

    data = response.json()

    if data["status"] == "success":
        return data["output"][0]
    else:
        raise Exception(data.get("message"))

# Change from day to evening
evening_room = change_room_scenario(
    "https://example.com/daytime-room.jpg",
    "Evening ambiance with warm lamp lighting, cozy atmosphere",
    "your_api_key"
)

オブジェクトの削除

def remove_interior_object(room_image, object_to_remove, api_key):
    """Remove an object from an interior image.

    Args:
        room_image: URL of room photo
        object_to_remove: Description of object to remove
    """
    response = requests.post(
        "https://modelslab.com/api/v6/interior/object_removal",
        json={
            "key": api_key,
            "init_image": room_image,
            "object_name": object_to_remove
        }
    )

    data = response.json()

    if data["status"] == "success":
        return data["output"][0]
    else:
        raise Exception(data.get("message"))

# Remove furniture
cleaned = remove_interior_object(
    "https://example.com/cluttered-room.jpg",
    "old sofa",
    "your_api_key"
)

インテリアミキサー

def mix_interior_objects(room_image, object_image, placement_prompt, api_key):
    """Add objects from one image into another room.

    Args:
        room_image: URL of the target room
        object_image: URL of image containing object to add
        placement_prompt: Description of how to place the object
    """
    response = requests.post(
        "https://modelslab.com/api/v6/interior/interior_mixer",
        json={
            "key": api_key,
            "init_image": room_image,
            "object_image": object_image,
            "prompt": placement_prompt,
            "width": 512,
            "height": 512,
            "num_inference_steps": 8,
            "guidance_scale": 7.5
        }
    )

    data = response.json()

    if data["status"] == "success":
        return data["output"][0]
    else:
        raise Exception(data.get("message"))

# Add furniture from another image
mixed = mix_interior_objects(
    "https://example.com/empty-room.jpg",
    "https://example.com/furniture.jpg",
    "Place the chair in the corner near the window",
    "your_api_key"
)

主なパラメーター

パラメーター	説明	値
init_image	部屋／建物の画像	画像 URL
prompt	デザインの説明	詳細なテキスト
negative_prompt	避けたい要素	"cluttered, low quality"
strength	変換の強度	0.0-1.0（通常は 0.7）
num_inference_steps	品質レベル	21、31、または 41
guidance_scale	プロンプトへの準拠度	1-20（通常は 7.5）
specific_object	必須アイテム	オブジェクト名
object_name	削除するオブジェクト	説明

ベストプラクティス

1. 詳細なデザインプロンプトを書く

✗ Bad: "modern room"
✓ Good: "Modern minimalist living room with Scandinavian furniture, white walls, oak floor, large windows, indoor plants"

Include: Style, furniture, colors, lighting, materials, atmosphere

2. 適切な強度の値を使用する

# Subtle changes
strength = 0.5

# Moderate redesign
strength = 0.7

# Complete transformation
strength = 0.9

3. 品質と速度のバランス

# Fast (21 steps)
num_inference_steps = 21

# Balanced (31 steps) - Recommended
num_inference_steps = 31

# Best quality (41 steps)
num_inference_steps = 41

4. 高品質な入力画像を使用する

• 明るく撮影された部屋の写真
• 空間がはっきりと見える画像
• 歪みが最小限であること
• 高解像度が望ましい

よくあるユースケース

バーチャルステージング

def stage_empty_room(room_image, style, api_key):
    """Stage an empty room for real estate listing."""
    return decorate_room(
        room_image,
        f"{style} furnished room with modern furniture, well-lit, professional",
        api_key
    )

# Stage for listing
staged = stage_empty_room(
    "https://example.com/empty-apartment.jpg",
    "Modern luxury",
    api_key
)

デザインのバリエーション

def create_design_variations(room_image, styles, api_key):
    """Generate multiple design style variations."""
    variations = []

    for style in styles:
        variant = redesign_interior(
            room_image,
            f"{style} interior design style",
            api_key
        )
        variations.append(variant)
        print(f"{style}: {variant}")

    return variations

# Generate variations
designs = create_design_variations(
    "https://example.com/room.jpg",
    ["Modern Scandinavian", "Industrial Loft", "Classic Traditional", "Bohemian"],
    api_key
)

リノベーション計画

def plan_renovation(current_room, desired_style, api_key):
    """Plan room renovation with before/after."""
    before = current_room

    after = redesign_interior(
        before,
        f"Renovated {desired_style} room with updated fixtures and furniture",
        api_key
    )

    return {"before": before, "after": after}

# Plan kitchen renovation
plan = plan_renovation(
    "https://example.com/old-kitchen.jpg",
    "modern farmhouse kitchen",
    api_key
)

部屋の全面改装

def complete_room_makeover(room_image, api_key):
    """Full room transformation workflow."""

    # Step 1: Remove unwanted items
    cleaned = remove_interior_object(
        room_image,
        "old furniture and clutter",
        api_key
    )

    # Step 2: Redesign space
    redesigned = redesign_interior(
        cleaned,
        "Modern minimalist interior with natural materials",
        api_key
    )

    # Step 3: Add specific decor
    final = decorate_room(
        redesigned,
        "Add cozy lighting and indoor plants",
        api_key,
        specific_object="pendant lamp"
    )

    return final

変更前／変更後のシナリオ

# Day to night transformation
day_room = "https://example.com/daytime.jpg"

night = change_room_scenario(
    day_room,
    "Evening ambiance with warm lighting, twilight outside windows",
    api_key
)

# Summer to winter
winter = change_room_scenario(
    day_room,
    "Winter scene with snow outside, cozy fireplace, warm interior",
    api_key
)

エラー処理

try:
    design = redesign_interior(room_image, prompt, api_key)
    print(f"Design created: {design}")
except Exception as e:
    print(f"Design generation failed: {e}")
    # Log error, try different prompt, notify user

パフォーマンスのヒント

1. 適切な推論ステップ数を使用する：31 ステップにすると、品質と速度のバランスが取れます
2. プロンプトを最適化する：明確で詳細なプロンプトが最適です
3. 類似したリクエストをまとめて処理する：複数のバリエーションをまとめて生成します
4. 結果をキャッシュする：生成したデザインを保存します
5. 品質を確認する：必要に応じて strength と guidance_scale を調整します

エンタープライズ API

専用リソースについては、以下をご覧ください。

# Enterprise endpoints
url = "https://modelslab.com/api/v1/enterprise/interior/interior"
url = "https://modelslab.com/api/v1/enterprise/interior/room_decorator"

リソース

• インテリア API ドキュメント：https://docs.modelslab.com/interior-api/overview
• インテリアデザイン：https://docs.modelslab.com/interior-api/interior
• ルームデコレーター：https://docs.modelslab.com/interior-api/room-decorator
• 間取り図の作成：https://docs.modelslab.com/interior-api/floor-planning
• API キーを取得：https://modelslab.com/dashboard

関連スキル

• modelslab-image-generation - 部屋の参考画像を生成する
• modelslab-image-editing - その他の編集ツール
• modelslab-sdk-usage - 公式 SDK を使用する