Research Article
Hybrid Deep Learning Algorithm-Based Food Recognition and Calorie Estimation
Table 4
Comparative analysis of the existing approaches with the proposed method.
| Evaluation metrics | Mask RCNN+YOLO V5 | Mask RCNN | YOLO | CNN |
| Accuracy | 0.97125 | 0.96225 | 0.95125 | 0.92125 | Sensitivity | 0.96754 | 0.95354 | 0.94754 | 0.93754 | Specificity | 0.96452 | 0.95152 | 0.94252 | 0.93452 | Precision | 0.96415 | 0.95015 | 0.94115 | 0.93415 | -score | 0.96584 | 0.95184 | 0.94433 | 0.93584 | MCC | 0.96241 | 0.95041 | 0.94341 | 0.93241 | NPV | 0.94592 | 0.93892 | 0.92692 | 0.91592 | FPR | 0.0154 | 0.0214 | 0.0294 | 0.0354 | FDR | 0.0148 | 0.0198 | 0.0238 | 0.0348 | FNR | 0.0161 | 0.0251 | 0.0321 | 0.0461 |
|
|