AI Paper Atlas

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

Image2Reverb: Cross-Modal Reverb Impulse Response Synthesis

Nikhil Singh, Jeff Mentch, Jerry Ng, Matthew Beveridge, Iddo Drori

ICCV 2021 · Vision-language and multimodal › Cross-modal learning · approach: Audio-visual source localization · 49 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
add depth maps as a fourth input channel (ablation)Image2Reverb · Test set T60 estimation (50 epochs)std of T60 error (%): 83.1 -> 78.8Improved
add depth maps as a fourth input channel (ablation)Image2Reverb · Test set T60 estimation (50 epochs)mean T60 error (%): -9.17 -> -6.03Improved
add differentiable T60 proxy loss term (ablation)Image2Reverb · Test set T60 estimation (50 epochs)std of T60 error (%): 85.97 -> 78.8Improved
add differentiable T60 proxy loss term (ablation)Image2Reverb · Test set T60 estimation (50 epochs)mean T60 error (%): -7.1 -> -6.03Improved
use Places365-pretrained ResNet50 encoder instead of random init (ablation)Image2Reverb · Test set T60 estimation (50 epochs)mean T60 error (%): 43.15 -> -6.03Improved
generate IR with the model instead of nearest-neighbor retrievalImage2Reverb · Test set T60 estimation (50 epochs)std of T60 error (%): 491.02 -> 78.8Improved
set constant depth map to 0 instead of estimated depth (ablation)Image2Reverb · Generated IR T60std of T60 (s): 1.54 -> 0.87Improved
set constant depth map to 0.5 instead of estimated depth (ablation)Image2Reverb · Generated IR T60mean T60 (s): 2.07 -> 3.62Improved
compare generated vs measured IRs on outdoor scenesImage2Reverb · Expert study (quality rating, Outdoor scenes)equivalence test p-value: .62Hurt
compare generated vs measured IRs on large scenesImage2Reverb · Expert study (quality rating, Large scenes)equivalence test p-value: <.001No 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