mirror of
https://github.com/blakeblackshear/frigate.git
synced 2025-10-27 10:52:11 +01:00
8a8a0c7dec
* Use cosine distance metric for vec tables * Only apply normalization to multi modal searches * Catch possible edge case in stddev calc * Use sigmoid function for normalization for multi modal searches only * Ensure we get model state on initial page load * Only save stats for multi modal searches and only use cosine similarity for image -> image search
55 lines
1.5 KiB
Python
55 lines
1.5 KiB
Python
"""Z-score normalization for search distance."""
|
|
|
|
import math
|
|
|
|
|
|
class ZScoreNormalization:
|
|
def __init__(self, scale_factor: float = 1.0, bias: float = 0.0):
|
|
"""Initialize with optional scaling and bias adjustments."""
|
|
"""scale_factor adjusts the magnitude of each score"""
|
|
"""bias will artificially shift the entire distribution upwards"""
|
|
self.n = 0
|
|
self.mean = 0
|
|
self.m2 = 0
|
|
self.scale_factor = scale_factor
|
|
self.bias = bias
|
|
|
|
@property
|
|
def variance(self):
|
|
return self.m2 / (self.n - 1) if self.n > 1 else 0.0
|
|
|
|
@property
|
|
def stddev(self):
|
|
return math.sqrt(self.variance) if self.variance > 0 else 0.0
|
|
|
|
def normalize(self, distances: list[float], save_stats: bool):
|
|
if save_stats:
|
|
self._update(distances)
|
|
if self.stddev == 0:
|
|
return distances
|
|
return [
|
|
(x - self.mean) / self.stddev * self.scale_factor + self.bias
|
|
for x in distances
|
|
]
|
|
|
|
def _update(self, distances: list[float]):
|
|
for x in distances:
|
|
self.n += 1
|
|
delta = x - self.mean
|
|
self.mean += delta / self.n
|
|
delta2 = x - self.mean
|
|
self.m2 += delta * delta2
|
|
|
|
def to_dict(self):
|
|
return {
|
|
"n": self.n,
|
|
"mean": self.mean,
|
|
"m2": self.m2,
|
|
}
|
|
|
|
def from_dict(self, data: dict):
|
|
self.n = data["n"]
|
|
self.mean = data["mean"]
|
|
self.m2 = data["m2"]
|
|
return self
|