Stochastic Depth
العمق العشوائي
تقنية تنظيم تُسقط طبقات كاملة عشوائياً أثناء التدريب، مما يُجبر الشبكة على المتانة تجاه تغيرات العمق.
A regularization technique that randomly drops entire layers during training, forcing the network to be robust to depth variations.
Also translated asالإسقاط العشوائي للطبقات
First appears in this corpus in: EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks (2019)
Appears in these papers
- A ConvNet for the 2020s2022in the sky ✦
- A ConvNet for the 2020s2022in the sky ✦
- Training Data-Efficient Image Transformers & Distillation Through Attention2021in the sky ✦
- Training Data-Efficient Image Transformers & Distillation Through Attention2021in the sky ✦
- DINOv2: Learning Robust Visual Features Without Supervision2023in the sky ✦
- EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks2019in the sky ✦