AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Pix2map: Cross-Modal Retrieval for Inferring Street Maps From Images

Xindi Wu, KwunFung Lau, Francesco Ferroni, Aljoša Ošep, Deva Ramanan

CVPR 2023 · Vision-language and multimodal › Cross-modal learning · approach: Unimodal-to-multimodal representation transfer · 6 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 cross-modal image-graph retrieval instead of unimodal nearest-neighbor retrieval (ablation)Pix2Map · Argoverse MapUpdateChamfer: 3.2168 -> 1.5908Improved
Add Chamfer-based partial credit to the contrastive training loss (ablation)Pix2Map · Argoverse MapUpdateChamfer: 3.2249 -> 1.7440Improved
Add edge loss to contrastive plus Chamfer training (ablation)Pix2Map · Argoverse MapUpdateChamfer: 1.7440 -> 1.5908Improved
Use early fusion of camera views instead of separate image encoders with average pooling (ablation)Pix2Map · Argoverse MapUpdateChamfer: 4.8129 -> 1.5908Improved
Resample graph vertices to be approximately equidistant (ablation)Pix2Map · Argoverse MapUpdateChamfer: 3.2663 -> 1.5908Improved
Restrict transformer attention using the graph adjacency mask (ablation)Pix2Map · Argoverse MapUpdateChamfer: 1.9309 -> 1.5908Improved
Add adjacency information as a node input feature (ablation)Pix2Map · Argoverse MapUpdateChamfer: 1.9241 -> 1.5908Improved
Add positional encodings to graph vertices (ablation)Pix2Map · Argoverse MapUpdateChamfer: 1.5908 -> 2.1564Hurt
Expand the retrieval library with unpaired maps (ablation)Pix2Map · Argoverse MapUpdateChamfer: 1.5908 -> 1.5208Improved
Increase Pittsburgh retrieval library size from 5.7k to 40k graphs (ablation)Pix2Map · Pittsburgh MapUpdateChamfer: 1.5908 -> 1.5208Improved
Use Pix2Map rather than TOPO-TRStreet-map inference · Argoverse MapUpdateChamfer: 3.0140 -> 2.0882Improved
Test on novel areas rather than previously seen regionsPix2Map · Pittsburgh MapExpand vs MapUpdateChamfer: 1.5908 -> 2.6654Hurt

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