English

LANA: Latency Aware Network Acceleration

Computer Vision and Pattern Recognition 2021-11-19 v2 Artificial Intelligence Machine Learning

Abstract

We introduce latency-aware network acceleration (LANA) - an approach that builds on neural architecture search techniques and teacher-student distillation to accelerate neural networks. LANA consists of two phases: in the first phase, it trains many alternative operations for every layer of the teacher network using layer-wise feature map distillation. In the second phase, it solves the combinatorial selection of efficient operations using a novel constrained integer linear optimization (ILP) approach. ILP brings unique properties as it (i) performs NAS within a few seconds to minutes, (ii) easily satisfies budget constraints, (iii) works on the layer-granularity, (iv) supports a huge search space O(10100)O(10^{100}), surpassing prior search approaches in efficacy and efficiency. In extensive experiments, we show that LANA yields efficient and accurate models constrained by a target latency budget, while being significantly faster than other techniques. We analyze three popular network architectures: EfficientNetV1, EfficientNetV2 and ResNeST, and achieve accuracy improvement for all models (up to 3.0%3.0\%) when compressing larger models to the latency level of smaller models. LANA achieves significant speed-ups (up to 5×5\times) with minor to no accuracy drop on GPU and CPU. The code will be shared soon.

Keywords

Cite

@article{arxiv.2107.10624,
  title  = {LANA: Latency Aware Network Acceleration},
  author = {Pavlo Molchanov and Jimmy Hall and Hongxu Yin and Jan Kautz and Nicolo Fusi and Arash Vahdat},
  journal= {arXiv preprint arXiv:2107.10624},
  year   = {2021}
}
R2 v1 2026-06-24T04:25:42.592Z