File size: 1,139 Bytes
aac350d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 | """
Pipeline package — local processing stages that run BEFORE the
orchestrator fans out to providers.
Flow:
raw user input (bytes / URL / base64)
↓
validation.py — InputValidator (reject bad input)
↓
preprocessing.py — ImagePreprocessor (decode, resize, normalize)
↓
hashing.py — ImageHasher (SHA-256 cache key)
↓
feature_extraction.py — FeatureExtractor (face crops, default-detector boxes)
↓
PipelineOutput (consumed by orchestrator)
The orchestrator receives a PipelineOutput, never raw user input.
"""
from pipeline.validation import InputValidator, ValidationResult
from pipeline.preprocessing import ImagePreprocessor, PreprocessedImage
from pipeline.hashing import ImageHasher
from pipeline.feature_extraction import FeatureExtractor, PipelineOutput
from pipeline.postprocessing import ResultPostprocessor
__all__ = [
"InputValidator",
"ValidationResult",
"ImagePreprocessor",
"PreprocessedImage",
"ImageHasher",
"FeatureExtractor",
"PipelineOutput",
"ResultPostprocessor",
]
|