Image segmentation basics in TensorFlow - Model Metrics & Evaluation

Metrics & Evaluation - Image segmentation basics

Which metric matters for Image Segmentation and WHY

In image segmentation, we want to know how well the model separates different parts of an image. The key metrics are Intersection over Union (IoU) and Dice coefficient. They measure how much the predicted area overlaps with the true area.

IoU is the size of the overlap divided by the size of the union of predicted and true areas. Dice is similar but gives more weight to overlap. These metrics help us understand if the model is correctly identifying the shapes and boundaries.

Confusion Matrix for Image Segmentation

For each pixel, the model predicts a class (like background or object). We can count:

      True Positive (TP): Pixel correctly predicted as object
      False Positive (FP): Pixel wrongly predicted as object
      True Negative (TN): Pixel correctly predicted as background
      False Negative (FN): Pixel wrongly predicted as object
    

Example confusion matrix for pixels:

      | Predicted Object | Predicted Background |
      |------------------|----------------------|
      | TP = 8000        | FN = 2000            |
      | FP = 1000        | TN = 9000            |
    

Precision vs Recall Tradeoff in Image Segmentation

Precision tells us how many predicted object pixels are actually correct. High precision means few false alarms.

Recall tells us how many true object pixels were found. High recall means few missed parts.

For example, in medical image segmentation, missing a tumor (low recall) is worse than marking some healthy tissue as tumor (lower precision). So recall is more important.

In other cases, like segmenting roads in maps, precision might be more important to avoid false roads.

What Good vs Bad Metric Values Look Like

Good segmentation models have IoU and Dice scores close to 1.0 (like 0.8 or above).

Bad models have low scores (below 0.5), meaning poor overlap.

Example:

• Good: IoU = 0.85, Dice = 0.90
• Bad: IoU = 0.40, Dice = 0.50

Also, balanced precision and recall above 0.8 is a sign of a reliable model.

Common Pitfalls in Image Segmentation Metrics

• Accuracy paradox: If most pixels are background, a model predicting all background can have high accuracy but no useful segmentation.
• Data leakage: Testing on images very similar to training can give falsely high metrics.
• Overfitting: Very high training IoU but low test IoU means the model memorizes training images, not generalizing.
• Ignoring class imbalance: Small objects can be missed if metrics focus on large background areas.

Self Check: Your model has 98% accuracy but 12% recall on object pixels. Is it good?

No, it is not good. The high accuracy is misleading because most pixels are background. The very low recall means the model misses most object pixels, so it fails to segment the important parts.

Key Result

IoU and Dice scores near 1.0 show good segmentation; high accuracy alone can be misleading due to class imbalance.

Practice

1. What is the main goal of image segmentation in TensorFlow?

easy

A. To compress images without losing quality

B. To label each pixel in an image with a class

C. To generate random noise images

D. To classify the entire image into one category

Solution

1. Step 1: Understand image segmentation purpose

   Image segmentation assigns a class label to every pixel, dividing the image into meaningful parts.

2. Step 2: Compare with other tasks

   Unlike classification, which labels the whole image, segmentation works at pixel level.

3. Final Answer:

   To label each pixel in an image with a class -> Option B

4. Quick Check:

   Image segmentation = pixel-wise labeling [OK]

Hint: Segmentation means labeling pixels, not whole images [OK]

Common Mistakes:

• Confusing segmentation with image classification
• Thinking segmentation compresses images
• Believing segmentation creates random images

2. Which TensorFlow layer is commonly used to extract features for image segmentation?

easy

A. Convolutional layer

B. Dense layer

C. Dropout layer

D. Flatten layer

Solution

1. Step 1: Identify feature extraction layers

   Convolutional layers scan images to find patterns and features useful for segmentation.

2. Step 2: Understand other layers' roles

   Dense layers are for classification, dropout prevents overfitting, flatten reshapes data but doesn't extract features.

3. Final Answer:

   Convolutional layer -> Option A

4. Quick Check:

   Feature extraction = convolutional layers [OK]

Hint: Convolutional layers find image features [OK]

Common Mistakes:

• Choosing dense layers for feature extraction
• Confusing flatten with convolution
• Ignoring convolution's role in images

3. What will be the shape of the output tensor after applying a softmax activation on a segmentation model predicting 3 classes for a 128x128 image?

medium

A. (128, 128, 3)

B. (3, 128, 128)

C. (128, 3, 128)

D. (128, 128)

Solution

1. Step 1: Understand output shape for segmentation

   Each pixel has a probability for each class, so output shape is height x width x classes.

2. Step 2: Apply to given image and classes

   Image size is 128x128, classes are 3, so output shape is (128, 128, 3).

3. Final Answer:

   (128, 128, 3) -> Option A

4. Quick Check:

   Output shape = (height, width, classes) [OK]

Hint: Output shape = image size + number of classes [OK]

Common Mistakes:

• Mixing class dimension position
• Forgetting classes dimension
• Using 2D shape instead of 3D

4. Given this TensorFlow code snippet for segmentation, what is the error?

model = tf.keras.Sequential([
  tf.keras.layers.Conv2D(16, 3, activation='relu', input_shape=(64, 64, 3)),
  tf.keras.layers.Dense(3, activation='softmax')
])

medium

A. Conv2D kernel size must be 1

B. Input shape is incorrect for images

C. Activation 'relu' is invalid

D. Dense layer cannot follow Conv2D without flattening

Solution

1. Step 1: Check layer sequence

   Conv2D outputs 4D tensor, Dense expects 2D input, so flattening is needed before Dense.

2. Step 2: Validate other options

   Input shape is valid for images, 'relu' is a valid activation, kernel size 3 is allowed.

3. Final Answer:

   Dense layer cannot follow Conv2D without flattening -> Option D

4. Quick Check:

   Dense after Conv2D needs flattening [OK]

Hint: Dense layers need flat input, add Flatten after Conv2D [OK]

Common Mistakes:

• Ignoring tensor shape mismatch
• Thinking kernel size must be 1
• Misunderstanding activation functions

5. You want to train a simple segmentation model in TensorFlow. Which data do you need to provide during training?

hard

A. Only images without labels

B. Images and image-level class labels

C. Images and their pixel-level masks

D. Random noise images

Solution

1. Step 1: Identify training data for segmentation

   Segmentation requires images and masks labeling each pixel's class to learn pixel-wise predictions.

2. Step 2: Compare with other data types

   Image-level labels are for classification, no labels means no supervision, noise images are irrelevant.

3. Final Answer:

   Images and their pixel-level masks -> Option C

4. Quick Check:

   Training segmentation = images + pixel masks [OK]

Hint: Segmentation training needs pixel masks, not just labels [OK]

Common Mistakes:

• Using only image-level labels
• Training without masks
• Confusing segmentation with classification data