computer-vision

name	computer-vision
description	Computer vision workflows — image data characterization, preprocessing and augmentation, architecture selection (CNN vs ViT), and evaluation metrics (mAP, IoU, FID, SSIM). Use when working with image or video data.
allowed_agents	["data","experiment"]

Computer Vision

Overview

Computer vision workflows require careful attention at every stage: understanding dataset characteristics first, building a sound preprocessing and augmentation pipeline, selecting an architecture matched to dataset size and task, and evaluating with task-appropriate metrics. This skill covers the full pipeline from raw images to model evaluation.

When to Use This Skill

Use this skill when:

• Working with image or video datasets (classification, detection, segmentation, generation)
• Designing or debugging a preprocessing/augmentation pipeline
• Selecting a model architecture for a vision task
• Computing vision-specific metrics (mAP, IoU, FID, SSIM, LPIPS)
• Transfer learning decisions (freeze vs. fine-tune, learning rate schedule)

Run the EDA skill first to understand file formats, directory structure, and basic counts. Use this skill for vision-specific analysis.

---

Image Dataset Characterization

Before writing any training code, profile your dataset thoroughly.

from pathlib import Path
from PIL import Image
import numpy as np
import matplotlib.pyplot as plt
from collections import Counter
import cv2

def characterize_dataset(image_dir, extensions=('.jpg', '.jpeg', '.png', '.tiff', '.bmp')):
    image_paths = [p for p in Path(image_dir).rglob('*') if p.suffix.lower() in extensions]
    print(f"Total images: {len(image_paths)}")

    widths, heights, channels_list, aspect_ratios = [], [], [], []
    channel_means, channel_stds = [], []

    for path in image_paths:
        with Image.open(path) as img:
            w, h = img.size
            c = len(img.getbands())
            widths.append(w)
            heights.append(h)
            channels_list.append(c)
            aspect_ratios.append(w / h)

            # Per-image channel stats (sample every Nth image to stay fast)
            if len(channel_means) < 500:
                arr = np.array(img.convert('RGB'), dtype=np.float32) / 255.0
                channel_means.append(arr.mean(axis=(0,1)))
                channel_stds.append(arr.std(axis=(0,1)))

    print(f"\nWidth  — min: {min(widths)}, max: {max(widths)}, mean: {np.mean(widths):.0f}")
    print(f"Height — min: {min(heights)}, max: {max(heights)}, mean: {np.mean(heights):.0f}")
    print(f"Aspect ratio — min: {min(aspect_ratios):.2f}, max: {max(aspect_ratios):.2f}, "
          f"mean: {np.mean(aspect_ratios):.2f}")
    print(f"Channels: {Counter(channels_list)}")

    means = np.array(channel_means).mean(axis=0)
    stds  = np.array(channel_stds).mean(axis=0)
    print(f"\nChannel means (RGB): {means.round(4)}")
    print(f"Channel stds  (RGB): {stds.round(4)}")

    # Class distribution (assumes ImageFolder structure: dir/class/image.jpg)
    classes = [p.parent.name for p in image_paths]
    class_counts = Counter(classes)
    print(f"\nClass distribution ({len(class_counts)} classes):")
    for cls, cnt in sorted(class_counts.items(), key=lambda x: -x[1]):
        print(f"  {cls}: {cnt}")

    # Plot size scatter
    fig, axes = plt.subplots(1, 2, figsize=(12, 4))
    axes[0].scatter(widths, heights, alpha=0.2, s=5)
    axes[0].set_xlabel('Width'); axes[0].set_ylabel('Height')
    axes[0].set_title('Image size distribution')
    axes[1].hist(aspect_ratios, bins=50)
    axes[1].set_xlabel('Aspect ratio (W/H)'); axes[1].set_title('Aspect ratio distribution')
    plt.tight_layout(); plt.savefig('dataset_profile.png', dpi=120)

    return {'widths': widths, 'heights': heights, 'means': means, 'stds': stds}

Key things to flag:

• Highly variable sizes: need a consistent resize strategy
• Extreme aspect ratios: letterbox or tile instead of naive resize
• Class imbalance >5:1: use weighted sampling or focal loss
• Few channels (grayscale dataset fed to RGB model): replicate channel

---

Preprocessing Pipeline

Resize Strategy by Task

import torchvision.transforms as T
import torchvision.transforms.functional as TF
import torch

# --- Classification: resize then center crop ---
transform_cls = T.Compose([
    T.Resize(256),          # Shorter side to 256
    T.CenterCrop(224),      # Crop to model input size
    T.ToTensor(),
    T.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
])

# --- Detection: letterbox (maintain aspect ratio, pad) ---
def letterbox(img, target_size=640, fill=(114, 114, 114)):
    """Resize with padding — preserves bounding box coordinates."""
    from PIL import Image
    w, h = img.size
    scale = target_size / max(w, h)
    new_w, new_h = int(w * scale), int(h * scale)
    img = img.resize((new_w, new_h), Image.BILINEAR)
    new_img = Image.new('RGB', (target_size, target_size), fill)
    pad_x = (target_size - new_w) // 2
    pad_y = (target_size - new_h) // 2
    new_img.paste(img, (pad_x, pad_y))
    return new_img, scale, pad_x, pad_y

# --- Segmentation: resize shortest side, then random crop ---
transform_seg = T.Compose([
    T.Resize(512, interpolation=T.InterpolationMode.BILINEAR),
    T.CenterCrop(512),
    T.ToTensor(),
    T.Normalize(mean=[0.485, 0.456, 0.406], std=[, , ])
])

Normalization

# ImageNet stats — use when:
#   - Fine-tuning a pretrained ImageNet model
#   - Dataset is natural images (photos, not microscopy, X-rays, etc.)
imagenet_mean = [0.485, 0.456, 0.406]
imagenet_std  = [0.229, 0.224, 0.225]

# Dataset-specific stats — use when:
#   - Training from scratch
#   - Domain-specific images (medical, satellite, microscopy)
#   - ImageNet stats produce bad convergence
def compute_dataset_stats(loader):
    mean = torch.zeros(3)
    std  = torch.zeros(3)
    n    = 0
    for imgs, _ in loader:
        # imgs: [B, C, H, W], values in [0,1]
        mean += imgs.mean(dim=[0, 2, 3]) * imgs.shape[0]
        std  += imgs.std(dim=[0, 2, 3])  * imgs.shape[0]
        n    += imgs.shape[0]
    return (mean / n).tolist(), (std / n).tolist()

normalize = T.Normalize(mean=imagenet_mean, std=imagenet_std)

Color Space Considerations

import cv2

# OpenCV reads BGR — always convert to RGB before passing to PyTorch
img_bgr = cv2.imread('image.jpg')
img_rgb = cv2.cvtColor(img_bgr, cv2.COLOR_BGR2RGB)

# Grayscale — replicate to 3 channels for RGB-pretrained models
gray = cv2.imread('xray.png', cv2.IMREAD_GRAYSCALE)
gray_3ch = cv2.cvtColor(gray, cv2.COLOR_GRAY2RGB)

# HSV — useful for color-based segmentation (e.g., plant health, traffic signs)
hsv = cv2.cvtColor(img_rgb, cv2.COLOR_RGB2HSV)
# H: 0–179, S: 0–255, V: 0–255 in OpenCV

---

Augmentation Strategies

By Task

Safe augmentations (valid for almost all tasks): horizontal flip, random crop, color jitter (brightness/contrast/saturation), Gaussian blur.

Risky augmentations — check task-specific constraints:

• Heavy rotation: wrong if orientation is discriminative (e.g., text OCR, medical "up" orientation)
• Aggressive color jitter: wrong if color is diagnostic (e.g., plant disease, pathology staining)
• Vertical flip: wrong for aerial/satellite imagery with meaningful up/down

import torchvision.transforms as T
import albumentations as A
from albumentations.pytorch import ToTensorV2

# --- Classification (torchvision) ---
train_transforms_cls = T.Compose([
    T.RandomHorizontalFlip(p=0.5),
    T.RandomResizedCrop(224, scale=(0.7, 1.0)),
    T.ColorJitter(brightness=0.3, contrast=0.3, saturation=0.3, hue=0.1),
    T.RandomGrayscale(p=0.05),
    T.RandomRotation(15),
    T.ToTensor(),
    T.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
])

# --- Detection (albumentations — transforms bboxes automatically) ---
train_transforms_det = A.Compose([
    A.HorizontalFlip(p=0.5),
    A.RandomResizedCrop(height=640, width=640, scale=(0.7, 1.0)),
    A.ColorJitter(brightness=0.3, contrast=0.3, saturation=0.3, hue=0.1, p=0.5),
    A.GaussianBlur(blur_limit=(3, 7), p=0.2),
    A.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
    ToTensorV2()
], bbox_params=A.BboxParams(format='yolo', label_fields=['class_labels']))


train_transforms_seg = A.Compose([
    A.HorizontalFlip(p=),
    A.RandomResizedCrop(height=, width=, scale=(, )),
    A.ElasticTransform(p=),
    A.GridDistortion(p=),
    A.Normalize(mean=[, , ], std=[, , ]),
    ToTensorV2()
])

Advanced Augmentations

# Mixup — blend two images and their labels linearly
def mixup_data(x, y, alpha=0.4):
    lam = np.random.beta(alpha, alpha)
    idx = torch.randperm(x.size(0))
    mixed_x = lam * x + (1 - lam) * x[idx]
    y_a, y_b = y, y[idx]
    return mixed_x, y_a, y_b, lam

def mixup_criterion(criterion, pred, y_a, y_b, lam):
    return lam * criterion(pred, y_a) + (1 - lam) * criterion(pred, y_b)

# CutMix — paste a patch from one image into another
def cutmix_data(x, y, alpha=1.0):
    lam = np.random.beta(alpha, alpha)
    idx = torch.randperm(x.size(0))
    _, _, H, W = x.shape
    cut_rat = np.sqrt(1 - lam)
    cut_w, cut_h = int(W * cut_rat), int(H * cut_rat)
    cx, cy = np.random.randint(W), np.random.randint(H)
    x1 = max(cx - cut_w // 2, 0); x2 = min(cx + cut_w // 2, W)
    y1 = max(cy - cut_h // 2, 0); y2 = min(cy + cut_h // 2, H)
    x[:, :, y1:y2, x1:x2] = x[idx, :, y1:y2, x1:x2]
    lam = 1 - (x2 - x1) * (y2 - y1) / (W * H)
    return x, y, y[idx], lam

# AutoAugment and AugMix via torchvision
auto_aug   = T.AutoAugment(policy=T.AutoAugmentPolicy.IMAGENET)
augmix     = T.AugMix()
rand_aug   = T.RandAugment(num_ops=2, magnitude=9)
trivial_aug = T.TrivialAugmentWide()  # strong, simple, no tuning needed

---

Architecture Selection

Decision Guide

Scenario	Recommendation	Library
Small dataset (<10k images)	Fine-tune ResNet50 or EfficientNet-B3 (pretrained on ImageNet)	torchvision.models
Medium dataset (10k–100k)	ViT-S/16 or ViT-B/16 with pretrained weights	timm
Large dataset (>100k)	ConvNeXt-Base or ViT-B pretrained, fine-tune all	timm
Speed-critical inference	MobileNetV3-Large, EfficientNet-B0	torchvision.models
Best accuracy, no speed constraint	ConvNeXt-XL, ViT-L, CLIP ViT-L/14	timm, OpenAI
Object detection	YOLOv8 (ultralytics), DINO, Faster R-CNN	ultralytics, torchvision
Semantic segmentation	SegFormer, DeepLabV3+	transformers, torchvision

import timm
import torchvision.models as models

# Pretrained ResNet50 — fine-tune for custom classification
model = models.resnet50(weights=models.ResNet50_Weights.IMAGENET1K_V2)
model.fc = torch.nn.Linear(model.fc.in_features, num_classes)  # replace head

# timm — massive model zoo with pretrained weights
model = timm.create_model('convnext_base', pretrained=True, num_classes=num_classes)
model = timm.create_model('vit_base_patch16_224', pretrained=True, num_classes=num_classes)
model = timm.create_model('efficientnet_b3', pretrained=True, num_classes=num_classes)

# List available models
timm.list_models('efficientnet*', pretrained=True)[:10]

---

Transfer Learning Guide

Freeze Backbone vs Fine-Tune All

Situation	Strategy
Very small dataset (<1k images), similar to ImageNet	Freeze backbone, train head only
Small dataset, different domain (medical, satellite)	Freeze early layers, fine-tune last 2–3 blocks + head
Medium+ dataset	Fine-tune all layers with discriminative LR
Large dataset, abundant compute	Train from scratch (still use pretrained init)

# Strategy 1: Freeze backbone, train head only
for param in model.parameters():
    param.requires_grad = False
for param in model.fc.parameters():   # or model.head, model.classifier
    param.requires_grad = True
optimizer = torch.optim.Adam(filter(lambda p: p.requires_grad, model.parameters()), lr=1e-3)

# Strategy 2: Discriminative learning rates (different LR per layer group)
optimizer = torch.optim.Adam([
    {'params': model.layer1.parameters(), 'lr': 1e-5},
    {'params': model.layer2.parameters(), 'lr': 1e-5},
    {'params': model.layer3.parameters(), 'lr': 1e-4},
    {'params': model.layer4.parameters(), 'lr': 1e-4},
    {'params': model.fc.parameters(),     'lr': 1e-3},
])

# Learning rate schedule: warmup + cosine decay
from torch.optim.lr_scheduler import OneCycleLR, CosineAnnealingLR

scheduler = OneCycleLR(
    optimizer,
    max_lr=1e-3,
    steps_per_epoch=len(train_loader),
    epochs=num_epochs,
    pct_start=0.05  # 5% warmup
)

---

CV Evaluation Metrics

Classification

from sklearn.metrics import accuracy_score, top_k_accuracy_score

# top-1 accuracy
top1 = accuracy_score(y_true, y_pred_classes)

# top-5 accuracy (multi-class problems with many classes)
top5 = top_k_accuracy_score(y_true, y_prob, k=5)
print(f"Top-1: {top1:.4f}, Top-5: {top5:.4f}")

Detection: IoU, mAP

import torchvision.ops as ops
import torch

def compute_iou(box1, box2):
    """
    box format: [x1, y1, x2, y2]
    Returns IoU scalar.
    """
    b1 = torch.tensor(box1, dtype=torch.float).unsqueeze(0)
    b2 = torch.tensor(box2, dtype=torch.float).unsqueeze(0)
    return ops.box_iou(b1, b2).item()

# mAP with torchmetrics (recommended)
from torchmetrics.detection.mean_ap import MeanAveragePrecision

metric = MeanAveragePrecision(iou_type='bbox')

# preds: list of dicts with keys 'boxes', 'scores', 'labels'
# targets: list of dicts with keys 'boxes', 'labels'
metric.update(preds, targets)
result = metric.compute()
print(f"mAP@0.5:     {result['map_50']:.4f}")
print(f"mAP@0.5:0.95:{result['map']:.4f}")

IoU threshold explanation:

• IoU@0.5 (PASCAL VOC): a detection is correct if the box overlaps ground truth by ≥50%
• mAP@0.5:0.95 (COCO): average mAP over thresholds 0.5, 0.55, …, 0.95 — stricter, preferred for modern benchmarks

Segmentation: mIoU, Dice

def compute_miou(pred_mask, true_mask, num_classes):
    """pred_mask, true_mask: integer class labels, shape [H, W]"""
    ious = []
    for cls in range(num_classes):
        pred_cls = (pred_mask == cls)
        true_cls = (true_mask == cls)
        intersection = (pred_cls & true_cls).sum()
        union = (pred_cls | true_cls).sum()
        if union == 0:
            continue  # class not present in either — skip
        ious.append(intersection / union)
    return sum(ious) / len(ious) if ious else 0.0

def dice_coefficient(pred_mask, true_mask, smooth=1e-6):
    intersection = (pred_mask * true_mask).sum()
    return (2 * intersection + smooth) / (pred_mask.sum() + true_mask.sum() + smooth)

Generation / Synthesis: FID, SSIM, LPIPS

# SSIM — structural similarity, higher is better (max 1.0)
from skimage.metrics import structural_similarity as ssim
import cv2

img1 = cv2.imread('original.png', cv2.IMREAD_GRAYSCALE)
img2 = cv2.imread('reconstructed.png', cv2.IMREAD_GRAYSCALE)
score = ssim(img1, img2, data_range=255)
print(f"SSIM: {score:.4f}")

# FID — Frechet Inception Distance (lower is better)
# Install: pip install pytorch-fid
# Usage from command line:
#   python -m pytorch_fid path/to/real_images path/to/generated_images

# LPIPS — perceptual similarity (lower is better)
# Install: pip install lpips
import lpips
import torch

loss_fn = lpips.LPIPS(net='alex')  # or 'vgg'
img1_t = torch.from_numpy(img1_rgb).permute(2,0,1).unsqueeze(0).float() / 127.5 - 1
img2_t = torch.from_numpy(img2_rgb).permute(2,0,1).unsqueeze(0).float() / 127.5 - 1
d = loss_fn(img1_t, img2_t)
print(f"LPIPS: {d.item():.4f}")

Metric summary:

Metric	Task	Better when
Top-1 / Top-5 accuracy	Classification	Higher
mAP@0.5, mAP@0.5:0.95	Detection	Higher
mIoU	Segmentation	Higher
Dice	Segmentation	Higher
FID	Generation	Lower
SSIM	Reconstruction	Higher
LPIPS	Perceptual quality	Lower

---

Best Practices

1. Profile the dataset before writing training code — size variance and class imbalance dictate key pipeline choices
2. OpenCV reads BGR — always convert to RGB before any DL library
3. Use albumentations for detection/segmentation — spatial transforms are applied consistently to boxes and masks
4. Validate augmentation visually — render 10 augmented samples before training to catch errors
5. Match normalization to pretraining — ImageNet stats for ImageNet-pretrained, dataset-specific otherwise
6. Monitor training loss AND sample predictions — loss curves alone miss silent failures (mode collapse, label errors)
7. Use mAP@0.5:0.95 for detection benchmarks — mAP@0.5 alone is too lenient for modern standards

---

Common Pitfalls

• Augmenting validation data: Apply only resize + normalize to val/test; augmentations go on train only
• Wrong normalization order: Normalize AFTER ToTensor (images in [0,1]) not before
• Not handling EXIF rotation: PIL respects EXIF by default; OpenCV does not — check orientation
• FID computed on too few images: FID is unreliable below ~10k samples; use at least 10k real + 10k generated
• Comparing mAP across IoU thresholds without clarifying: Always state which threshold you report