菜单
Resize images preserving aspect ratio then zero-pad to a square to avoid distortion artifacts in face crops or object detection inputs
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Memory-efficient Keras generator that streams sharded CSV files in chunks, renders strokes to images on-the-fly, and yields batches for training on datasets too large for memory
Stack 1D convolutions for local feature extraction before bidirectional LSTMs to classify variable-length stroke sequences into hundreds of doodle categories
Partition a large dataset into N balanced shards using integer key modulo arithmetic for reproducible, class-interleaved splits across CSV files
Normalize raw stroke coordinates to 0-255 range, resample at uniform arc-length spacing, then apply Ramer-Douglas-Peucker simplification
Render stroke sequences to grayscale images with temporal intensity encoding where earlier strokes are brighter and later strokes fade to encode drawing order
Score each face in a video as anomalous by computing L2 distance from the embedding centroid of all faces, then convert to probability via logistic function
| name | cv-isotropic-resize-with-padding |
| description | Resize images preserving aspect ratio then zero-pad to a square to avoid distortion artifacts in face crops or object detection inputs |
Naively resizing a rectangular image to a square distorts aspect ratio, creating artifacts that confuse classifiers (especially for faces). Isotropic resize scales the image so the longer side matches the target size, then zero-pads the shorter side. This preserves proportions while producing a fixed-size square input for CNNs.
import cv2
import numpy as np
def isotropic_resize(img, size, interpolation=cv2.INTER_AREA):
h, w = img.shape[:2]
if w > h:
new_w = size
new_h = int(h * size / w)
else:
new_h = size
new_w = int(w * size / h)
resized = cv2.resize(img, (new_w, new_h), interpolation=interpolation)
# Zero-pad to square
canvas = np.zeros((size, size, 3), dtype=np.uint8)
canvas[:new_h, :new_w] = resized
return canvas
face_crop = isotropic_resize(face_crop, 224)
INTER_AREA for downsampling (anti-aliased), INTER_LINEAR for upsampling