Modifying YOLOv8 algorithm by adding ReLU function from the backbone to improve the detection of small objects

[claysha]

Is adding ReLU to YOLOv8 a good idea knowing that YOLOv8 already uses the SILU function. if yes, what could be the impact in terms of improving the algorithm