Overview

SciPy 1.16, released on June 22, 2025, adds 2D FIR filters and extends Array API support to the signal and stats modules.

Main Features

2D FIR filters

The scipy.signal module adds support for 2D FIR (Finite Impulse Response) filters for image processing.

python 
from scipy.signal import convolve2d
import numpy as np

# 2D FIR filter: smoothing kernel
image = np.random.rand(100, 100)
kernel = np.ones((3, 3)) / 9
smoothed = convolve2d(image, kernel, mode='same')
print(f'Shape: {smoothed.shape}')

Array API for signal and stats

The scipy.signal and scipy.stats modules now support the standard Array API, enabling backends like CuPy or JAX.

python 
from scipy import stats
import numpy as np

data = np.random.normal(0, 1, 1000)
result = stats.describe(data)
print(f'Mean: {result.mean:.3f}')
print(f'Variance: {result.variance:.3f}')

Sources