OFFLINE

Building efficient, SOTA AI models for companies

Participants

End Date

Ended

Total use case computation budget

0 F of 0 F utilized

Todays submits

0/5

Training Dataset ID

DwwCRMl5

CPU "- " Pod Limit

0 Cores

- GB

GPU "-" Pod Limit

- GPU

- GB VRAM

Accuracy (PCK):

Variable: {accuracy (PCK)}
Definition: The PCK (Percentage of Correct Keypoints) measures how accurately the model detects keypoints compared to the ground truth. A keypoint is considered correct if the distance between the predicted and ground truth keypoints is less than a defined threshold.
PCK Calculation:
- Threshold: Set at 0.2.
- Distance Calculation: Compute the Euclidean distance between the predicted keypoints and the ground truth keypoints using: distances=torch.linalg.norm(predicted_keypoints−ground_truth_keypoints,dim=2)
- Correct Keypoints: A keypoint is considered correct if the distance is less than the threshold. This is determined by: correct_keypoints=(distances<threshold).float()
- PCK Value: The PCK value is calculated as the mean of correct keypoints: pck=correct_keypoints.mean().item()
- Weight in the Score: 80% of the total score is based on this accuracy measure, making it the most significant factor in the final score.

Code Block

python

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

# Calculate PCK
def calculate_pck(predicted_keypoints, ground_truth_keypoints, threshold=0.2):
    # Calculate Euclidean distance between predicted and ground truth keypoints
    distances = torch.linalg.norm(predicted_keypoints - ground_truth_keypoints, dim=2)
    
    # Determine correct keypoints (distance < threshold)
    correct_keypoints = (distances < threshold).float()
    
    # Calculate PCK as mean of correct keypoints
    pck = correct_keypoints.mean().item()
    
    return pck

# Example usage
predicted = torch.rand(16, 17, 2)  # batch_size, num_keypoints, coordinates
ground_truth = torch.rand(16, 17, 2)
    
accuracy = calculate_pck(predicted, ground_truth)
print(f"PCK Accuracy: {accuracy:.4f}")