AI Paper Atlas

Research map / Architectures and efficiency / Neural architecture search

LANA: Latency Aware Network Acceleration

Pavlo Molchanov, Jimmy Hall, Hongxu Yin, Jan Kautz, Nicolo Fusi, Arash Vahdat

ECCV 2022 · Architectures and efficiency › Neural architecture search · approach: Latency-aware architecture search · 7 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use layer-wise instead of block-wise distillation for pretraining (ablation)LANA candidate pretraining · EfficientNetV1-B2 teacher mimickingtop-1 accuracy (%): 68.21 -> 76.52Improved
use layer-wise instead of block-wise distillation (ablation)LANA candidate pretraining · EfficientNetV2-B3 teacher mimickingtop-1 accuracy (%): 73.53 -> 76.52Improved
use layer-wise instead of block-wise distillation (ablation)LANA candidate pretraining · ResNeST-50 teacher mimickingtop-1 accuracy (%): 77.54 -> 80.61Improved
use ILP search instead of random sampling (ablation)LANA architecture search · 0.45xEfficientNetV1-B2top-1 accuracy (%): 76.44 -> 79.28Improved
use ILP instead of SNAS differentiable search (ablation)LANA architecture search · 0.45xEfficientNetV1-B2top-1 accuracy (%): 74.20 -> 79.28Improved
use ILP instead of E-NAS REINFORCE search (ablation)LANA architecture search · 0.45xEfficientNetV1-B2top-1 accuracy (%): 78.85 -> 79.28Improved
increase candidate pool from 2 to all 197 operations per layer (ablation)LANA search · 0.55xEfficientNetV1-B2top-1 accuracy (%): 79.40 -> 80.00Improved
increase candidate pool from 5 to 10 operations per layer (ablation)LANA search · 0.55xEfficientNetV1-B2top-1 accuracy (%): 79.52 -> 79.66Improved
load pretrained candidate weights instead of reinitializing before finetuning (ablation)LANA 0.45xB2 · ImageNet-1K valtop-1 accuracy (%): 79.42 -> 79.71Improved
add knowledge distillation during finetuning (ablation)LANA 0.45xB2 · ImageNet-1K valtop-1 accuracy (%): 0.65% degradation withoutImproved
use LANA 0.45xB2 instead of distilled vanilla EfficientNetV1-B0 (ablation)student architecture · ImageNet-1K valtop-1 accuracy (%): 78.72 -> 79.71Improved
add all operations beyond skip-only single-shot (ablation)LANA 0.45xEfficientNetV1-B2 · ImageNet-1K valtop-1 accuracy (%): 78.68 -> 79.71Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy