AI Paper Atlas

Research map / Architectures and efficiency / Neural architecture search

Dynamic Neural Network for Multi-Task Learning Searching Across Diverse Network Topologies

Wonhyeok Choi, Sunghoon Im

CVPR 2023 · 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 the searched task-adaptive network with M = 7 instead of AdaShareMulti-task learning · NYU-v2Δ_T: +8.9 vs +11.1Improved
Use the searched task-adaptive network with M = 9 instead of Cross-StitchMulti-task learning · CityscapesΔ_T: +2.8 vs +6.6Improved
Use the searched task-adaptive network with M = 9 instead of Cross-StitchMulti-task learning · PASCAL-ContextΔ_T: +0.1 vs +1.4Improved
Use the proposed network with M = 5 instead of Gumbel-Matrix20-way multi-task classification · OmniglotTest Acc. (%): 93.52 -> 95.71Improved
Add flow restriction (ablation)Ours (M = 7) · NYU-v2Δ_T: +11.0 -> +11.1No clear effect
Add read-in/out layers (ablation)Ours (M = 7) · NYU-v2Δ_T: +10.1 -> +11.1Improved
Add flow-based reduction (ablation)Ours (M = 7) · NYU-v2Δ_T: +11.5 -> +11.1Hurt
Add flow-based reduction (ablation)Ours (M = 7) · NYU-v2# of Param: 6.50 -> 1.31Improved
Add squeeze loss (ablation)Ours (M = 7) · NYU-v2Δ_T: +10.9 -> +11.1Improved
Add warm-up stage (ablation)Three-stage training · NYU-v2Δ_T: -4.3 -> 0.0Improved
Add search and flow-based reduction stages (ablation)Three-stage training · NYU-v2Δ_T: -6.1 -> 0.0Improved
Add fine-tuning stage (ablation)Three-stage training · NYU-v2Δ_T: -8.9 -> 0.0Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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