Some questions about gradient equation of C4 layer

[sihan-shao]

Hello author！In this paper, I am very inspired by your insights about the influence of different feature maps on small target detection, but I have a question about gradient equation in the paper.

In the gradient formula of C4, the backpropagation gradient information of P2 layer should pass through the fusion factor between P'2 and P'3. Why is it not reflected in the formula?

I am not very clear about this question or can you please provide appendix mentioned in this paper?

[Muqianli]

Hello author！In this paper, I am very inspired by your insights about the influence of different feature maps on small target detection, but I have a question about gradient equation in the paper.

In the gradient formula of C4, the backpropagation gradient information of P2 layer should pass through the fusion factor between P'2 and P'3. Why is it not reflected in the formula?

I am not very clear about this question or can you please provide appendix mentioned in this paper?

Hello! Thank you for your question! In this gradient equation, we use α34 and C4 as an example to analyze fusion factor in the backpropagation gradient. During the derivation, we omit α23 and α45 for convenience. The gradient of ∆C4 can be represent as two parts, from deep layers (P4 and P5) and from shallow layers (P2 and P3). In further research, we discover more interesting things about fusion factor, which will be presented in the follow-up work. Thank you very much!