M3-098M3: Deep Learning FoundationsConvolution & Vision FoundationsEasy
⚡Flashcard Drill➔Mastery:
Recursive Calculation of Receptive Field in Convolutional Networks
🎯 Active Recall Self-Check (Draft First, Verify After)
Login-free practice tracker: recall core answer, audit pitfalls, and log review status.
Step 1 · Draft your concise answer & assumptions
Step 2 · Check Pitfalls & Follow-up Inquiries
📐 Mathematical Definition
⚡ One-Sentence Core Answer
A feature's receptive field expands recursively via , where cumulative stride exponentially magnifies downsampled kernels.
📌 Key Takeaways
- •Recursive formula: , with base condition
- •Jump / Effective Stride: ; strides accumulate multiplicatively
- •Effective vs Theoretical: due to Gaussian-like gradient concentration, the effective receptive field covers only a fraction of the theoretical field
📐 Mathematical Derivations
Mathematical Formulations (Araujo et al., Distill 2019):
Let be the receptive field of layer , be kernel size, stride, and cumulative stride (jump).
- Base Initialization: .
- Recursive Step for Layer $l$:
,
.
Intuition: A kernel of size covers intervals. Each interval in layer corresponds to pixels in the original input image. Thus, each step expands the coverage by .
Theoretical vs Effective Receptive Field (ERF):
Luo et al. (NeurIPS 2016) proved that the distribution of impact of an input pixel on a central output feature decays as a 2D Gaussian. The Effective Receptive Field (ERF) (containing 95% of gradient energy) scales with , occupying only a central fraction () of the theoretical bounding box.
Let be the receptive field of layer , be kernel size, stride, and cumulative stride (jump).
- Base Initialization: .
- Recursive Step for Layer $l$:
,
.
Intuition: A kernel of size covers intervals. Each interval in layer corresponds to pixels in the original input image. Thus, each step expands the coverage by .
Theoretical vs Effective Receptive Field (ERF):
Luo et al. (NeurIPS 2016) proved that the distribution of impact of an input pixel on a central output feature decays as a 2D Gaussian. The Effective Receptive Field (ERF) (containing 95% of gradient energy) scales with , occupying only a central fraction () of the theoretical bounding box.
🏭 Production Trade-offs
Design implication for Dense Tasks: In semantic segmentation and object detection, theoretical receptive field must cover the entire input image (). Architectures achieve this via dilated convolutions (DeepLab) or Feature Pyramid Networks (FPN).
⚠️ Common Interview Pitfalls
- ✕Assuming every pixel inside the theoretical receptive field contributes equally; peripheral pixels contribute near-zero gradient signal
- ✕Forgetting that pooling layers act as kernels with and in receptive field recursion
🎯 Interviewer Follow-up Inquiries
- ?Why does the Effective Receptive Field (ERF) scale with rather than linearly with depth ?
- ?How do dilated convolutions scale the kernel parameter in the recursive receptive field formula?
⚡
Test muscle memory on this topic via 3-Layer Flashcards
Active Recall: Mentally formulate Prompt ➔ Math Derivation ➔ Benchmark Implementation in 3 minutes.
📚