The video scene segmentation operator uses a multimodal large model to perform shot/scene segmentation, global character recognition, scene-level character association, and character image extraction on the input video. The operator outputs scene summary results, a character registry, sliced videos for each scene, and image files archived by character, facilitating subsequent retrieval, editing, and content understanding.

Operator ID: las_video_scene_seg

Core capabilities #

• Supports scene segmentation based on VLM, and also supports equal-duration segmentation when min_segment_duration == max_segment_duration.
• Supports global character extraction and deduplication aggregation, generating a character registry.
• Supports character association within scenes, outputting the time intervals when characters appear in scenes, key frame timestamps, and bbox information.
• Supports automatic extraction of independent video files for each scene.
• Supports extracting and filtering representative images for each character, outputting them archived by character.
• Supports outputting token usage and LLM request counts to facilitate cost evaluation.

Effect demonstration #

The above are sample results for some outputs.

Supported regions #

• Beijing: cn-beijing
• Shanghai: cn-shanghai
• Guangzhou: cn-guangzhou

Operator performance #

Input requirements #

Output requirements #

• Billing standards

  Detailed items	Billing standards description
  Billing item	Billing is based on usage calculated from the duration of the input video.
  Billing type	Usage-based billing, unit: CNY/minute, billed hourly based on actual usage.
  Unit price	1.5 CNY/minute

• Billing details
  Billing formula: Total cost = unit price * usage

API usage instructions #

The API for this operator is an asynchronous task interface. You need to submit a task via the submit interface, then use the Poll interface to check the task status and obtain the operator processing result. Meanwhile, this operator supports asynchronous callback capability (Callback). You can set the callback parameter callback.url when initiating the operator task. LAS will send an HTTP POST notification to this URL. For guidance on using asynchronous callbacks, refer to Asynchronous callback API.

Submit #

Interface description #

Submitting an asynchronous task using the video scene segmentation operator supports scene segmentation, character extraction, scene slicing, and result upload.

Request parameters #

Response parameters #

Example #

Request example

# Please set OUTPUT_TOS_PATH to a tos path writable by this account
export OUTPUT_TOS_PATH="tos://path/to/scene_seg_results"

curl --location "https://operator.las.ap-southeast-1.volces.com/api/v1/submit" --header "Content-Type: application/json" --header "Authorization: Bearer $LAS_API_KEY" --data '{
    "operator_id": "las_video_scene_seg",
    "operator_version": "v1",
    "data": {
        "video_url": "https://las-ai-qa-online.tos-cn-beijing.volces.com/operator_cards_serving/public/qa/las_video_scene_seg/v1/sample_video.mp4",
        "output_tos_path": "$OUTPUT_TOS_PATH",
        "min_segment_duration": 4.0,
        "max_segment_duration": 10.0,
        "seg_mode": "precise"
    }
}'

Response example

{
    "metadata": {
        "task_id": "task-20260323153012-abc123",
        "task_status": "PENDING",
        "business_code": "0",
        "error_msg": ""
    }
}

Poll #

Interface description #

Submitting an asynchronous task using the video scene segmentation operator supports scene segmentation, character extraction, scene slicing, and result upload.

Request parameters #

Response parameters #

Example #

Request example

curl --location "https://operator.las.ap-southeast-1.volces.com/api/v1/poll" --header "Content-Type: application/json" --header "Authorization: Bearer $LAS_API_KEY" --data '{
    "operator_id": "las_video_scene_seg",
    "operator_version": "v1",
    "task_id": "task-20260323153012-abc123"
}'

Response example

{
    "metadata": {
        "task_id": "task-20260323153012-abc123",
        "task_status": "COMPLETED",
        "submit_time": "2026-07-21T16:04:06+08:00",
        "end_time": "2026-07-21T16:04:06+08:00",
        "business_code": "0",
        "error_msg": ""
    },
    "data": {
        "video_duration": 98.4,
        "segment_count": 12,
        "character_count": 5,
        "segments_url": "tos://path/to/scene_seg_results/9f4b7c3d12ab_20260323_153012/segments.json",
        "characters_registry_url": "tos://path/to/scene_seg_results/9f4b7c3d12ab_20260323_153012/characters_registry.json"
    }
}

Error code #