Implementation:NVIDIA NeMo Curator ClipFrameExtractionStage
Appearance
| Knowledge Sources | |
|---|---|
| Domains | Data_Curation, Video_Processing |
| Last Updated | 2026-02-14 17:00 GMT |
Overview
Concrete tool for extracting image frames from video clips at configurable frame rates provided by NeMo Curator.
Description
The ClipFrameExtractionStage processes video clips to extract image frames using configurable extraction policies and target frame rates. It supports multiple decoder backends and stores frames keyed by FrameExtractionSignature for downstream stages.
Usage
Import this stage after clipping stages and before captioning, embedding, or aesthetic filtering stages that require image frames.
Code Reference
Source Location
- Repository: NeMo Curator
- File: nemo_curator/stages/video/clipping/clip_frame_extraction.py
- Lines: L33-134
Signature
@dataclass
class ClipFrameExtractionStage(ProcessingStage[VideoTask, VideoTask]):
extraction_policies: tuple[FrameExtractionPolicy, ...] = (FrameExtractionPolicy.sequence,)
extract_purposes: list[FramePurpose] | None = None
target_res: tuple[int, int] | None = None
verbose: bool = False
num_cpus: int = 3
target_fps: list[float | int] | None = None
name: str = "clip_frame_extraction"
Import
from nemo_curator.stages.video.clipping.clip_frame_extraction import ClipFrameExtractionStage
I/O Contract
Inputs
| Name | Type | Required | Description |
|---|---|---|---|
| task | VideoTask | Yes | Video with clips having buffer data |
Outputs
| Name | Type | Description |
|---|---|---|
| task | VideoTask | Video with clip.extracted_frames dict keyed by FrameExtractionSignature |
Usage Examples
from nemo_curator.stages.video.clipping.clip_frame_extraction import ClipFrameExtractionStage
from nemo_curator.utils.decoder_utils import FrameExtractionPolicy
stage = ClipFrameExtractionStage(
extraction_policies=(FrameExtractionPolicy.sequence,),
target_fps=[2.0],
num_cpus=3,
)
Related Pages
Implements Principle
Page Connections
Double-click a node to navigate. Hold to expand connections.
Principle
Implementation
Heuristic
Environment