AI Paper Atlas

Research map / Video / Video segmentation and tracking

High Speed and High Dynamic Range Video with an Event Camera

Henri Rebecq, René Ranftl, Vladlen Koltun, Davide Scaramuzza

TPAMI 2019 · Video › Video segmentation and tracking · approach: Event-based video reconstruction · 795 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 learned recurrent network instead of HF and MRevent-to-video reconstruction · Event Camera DatasetSSIM: 0.5009 -> 0.6157Improved
use the learned network instead of HF and MRevent-to-video reconstruction · Event Camera DatasetLPIPS: 0.5173 -> 0.4092Improved
use the learned network instead of HF and MR on synthetic eventsevent-to-video reconstruction · Synthetic event sequencesLPIPS: 0.467 -> 0.302Improved
add the temporal loss (ablation)E2VID · Event Camera Datasettemporal error: 2.0810 -> 1.4257Improved
add the temporal loss (ablation)E2VID · Event Camera DatasetSSIM: 0.5935 -> 0.6157Improved
add the recurrent connection (ablation)E2VID · Event Camera Datasettemporal error: 3.9964 -> 1.4257Improved
add the recurrent connection (ablation)E2VID · Event Camera DatasetSSIM: 0.5364 -> 0.6157Improved
replace vanilla RNN with stacked ConvLSTM (ablation)E2VID · tau=5 ms windowsSSIM: 0.2685 -> 0.5800Improved
replace vanilla RNN with stacked ConvLSTM (ablation)E2VID · tau=50 ms windowsSSIM: 0.6100 -> 0.6157No clear effect
apply reconstructions to classification with a pretrained CNN instead of HATSobject classification · N-Caltech101accuracy: 0.700 -> 0.866Improved
apply reconstructions instead of HATS on a synthetic datasetobject classification · N-MNISTaccuracy: 0.991 -> 0.983Hurt
feed reconstructions to a CNN trained on image data without fine-tuningobject classification · N-Caltech101accuracy: 0.700 -> 0.821Improved

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