AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Deep Patch Visual Odometry

Zachary Teed, Lahav Lipson, Jia Deng

NeurIPS 2023 · Architectures and efficiency › RNNs and normalization · approach: Dataset condensation and flow estimation · 39 citations (OpenAlex)

Read the paper

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 sparse patch-based correspondence instead of dense flow (DPVO vs DROID-VO)DPVO · EuRoC MAVATE [m] avg: 0.186 -> 0.105Improved
use sparse patch-based correspondence instead of dense flow (DPVO vs DROID-VO)DPVO · TartanAir test splitATE avg: 0.58 -> 0.21Improved
use sparse patch-based correspondence instead of dense flow (DPVO vs DROID-VO)DPVO · TUM-RGBD freiburg1ATE avg: 0.098 -> 0.089Improved
use 96 patches / 10-frame window (Default) instead of 48 patches / 7-frame (Fast)DPVO · EuRoC MAVATE [m] avg: 0.129 -> 0.105Improved
use 96 patches / 10-frame window (Default) instead of Fast on ICL-NUIMDPVO · ICL-NUIMATE avg: 0.093 -> 0.097No clear effect
use patches (larger than 1x1) instead of point features (ablation)DPVO · TartanAir validationaccuracyImproved
add 1D temporal convolution to update operator (ablation)DPVO · TartanAir validationaccuracyImproved
add softmax aggregation (message passing) to update operator (ablation)DPVO · TartanAir validationaccuracyImproved
random patch centroids instead of SIFT/ORB/SuperPoint/high-gradient keypoints (ablation)DPVO · TartanAir validationaccuracyImproved
increase number of patches beyond 96 (ablation)DPVO · EuRoCATENo clear effect

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