Research map / Architectures and efficiency / RNNs and normalization
Learnable Visual Markers
Oleg Grinchuk, Vadim Lebedev, Victor Lempitsky
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Replace the learned synthesizer with a standard QR encoder | Learned marker recognizer · Marker recognition under rendered distortions | accuracy: 85% | Hurt |
| Replace both learned networks with a QR encoder and QR decoder | Learned marker system · Markers processed by the renderer at typical blur | decodability | Hurt |
| Reduce the number of maps in the recognizer's convolutional layers (ablation) | Learned marker system · Recognition accuracy across bit-string lengths | accuracy | Hurt |
| Reduce spatial variation in the rendering network to σ=0.05 (ablation) | Learned marker system · Recognition across bit-string lengths | code capacity | Improved |
| Increase marker size (ablation) | Learned marker system · Texture-loss marker generation | pattern diversity | Improved |
| Add texture loss to the learning objective (ablation) | Learned marker system · Stylized marker generation | code capacity | Hurt |
| Increase marker size (ablation) | Learned marker system · Recognition across marker sizes | code capacity | Mixed |
Closest papers by meaning
- DeepTag: A General Framework for Fiducial Marker Design and Detection (TPAMI 2022 · similarity 0.84)
- Texture Networks: Feed-forward Synthesis of Textures and Stylized Images (ICML 2016 · similarity 0.83)
- A Powerful Generative Model Using Random Weights for the Deep Image Representation (NeurIPS 2016 · similarity 0.83)
- Deep Virtual Markers for Articulated 3D Shapes (ICCV 2021 · similarity 0.83)
- Diversified Texture Synthesis With Feed-Forward Networks (CVPR 2017 · similarity 0.83)
- Learning to Manipulate Artistic Images (AAAI 2024 · similarity 0.81)
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.
