Table 5 The comparative analysis of baseline model with proposed model with different metrics.
From: Modified energy-based GAN for intensity in homogeneity correction in brain MR images
S.No | Metric | Baseline model | Proposed model | Comparison |
|---|---|---|---|---|
1 | FLOPs (Floating Point Operations) | High | Reduced | The proposed model should have optimized operations |
2 | Inference Time (ms) | Longer | Shorter | Optimized model should be faster |
3 | Memory Footprint (MB) | Large | Smaller | Efficient architectures reduce memory load |
4 | Power Consumption (W) | Higher (Approximately 5 to 6Joules per image) | Lower (approximately 0.8 to 1.5Â J per image) | The proposed model should use less energy per operation |
5 | Inference Time (on GPU) | 100Â ms | 25Â ms | Energy-efficient models reduce inference time |
