CVLGApr 29, 2019

Progressive Differentiable Architecture Search: Bridging the Depth Gap between Search and Evaluation

arXiv:1904.12760v1727 citationsHas Code
Originality Incremental advance
AI Analysis

This addresses a key bottleneck in neural architecture search for computer vision, enabling more efficient and accurate model discovery, though it is incremental in improving existing differentiable methods.

The paper tackles the accuracy drop in differentiable neural architecture search due to depth mismatch between search and evaluation by introducing a progressive depth growth algorithm, achieving state-of-the-art performance on CIFAR and ImageNet with reduced search time (~7 hours on a single GPU).

Recently, differentiable search methods have made major progress in reducing the computational costs of neural architecture search. However, these approaches often report lower accuracy in evaluating the searched architecture or transferring it to another dataset. This is arguably due to the large gap between the architecture depths in search and evaluation scenarios. In this paper, we present an efficient algorithm which allows the depth of searched architectures to grow gradually during the training procedure. This brings two issues, namely, heavier computational overheads and weaker search stability, which we solve using search space approximation and regularization, respectively. With a significantly reduced search time (~7 hours on a single GPU), our approach achieves state-of-the-art performance on both the proxy dataset (CIFAR10 or CIFAR100) and the target dataset (ImageNet). Code is available at https://github.com/chenxin061/pdarts.

Code Implementations4 repos
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes