AI Paper Atlas

Research map / Detection and segmentation / CNN features

Deep Learning for Light Field Saliency Detection

Tiantian Wang, Yongri Piao, Xiao Li, Lihe Zhang, Huchuan Lu

ICCV 2019 · Detection and segmentation › CNN features · approach: Domain-specific segmentation and detection · 125 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 focal stack stream to all-focus image (ablation)two-stream VGG19 network (late fusion + att + LSTM) · LFSDmaxF: 0.817 -> 0.863Improved
add focal stack stream to all-focus image (ablation)two-stream VGG19 network (late fusion + att + LSTM) · Our light field datasetmaxF: 0.819 -> 0.868Improved
early integration of focal slices instead of all-focus only (ablation)VGG19 stream · Our light field datasetmaxF: 0.819 -> 0.847Improved
layer-wise integration instead of all-focus only (ablation)VGG19 stream · Our light field datasetmaxF: 0.819 -> 0.823No clear effect
late integration by element-wise addition instead of all-focus only (ablation)VGG19 stream · Our light field datasetmaxF: 0.819 -> 0.822No clear effect
add attention without ConvLSTM to late fusion (ablation)late fusion (d) · Our light field datasetMAE: 0.110 -> 0.227Hurt
add ConvLSTM on top of attention (ablation)late fusion + att · Our light field datasetmaxF: 0.819 -> 0.863Improved
add ConvLSTM on top of attention (ablation)late fusion + att · LFSDMAE: 0.236 -> 0.100Improved
add adversarial examples to fine-tuning (ablation)att + LSTM model · LFSDmaxF: 0.851 -> 0.863Improved
add adversarial examples to fine-tuning (ablation)att + LSTM model · Our light field datasetMAE: 0.072 -> 0.070Improved
add adversarial examples to fine-tuning (ablation)att + LSTM model · Our light field datasetmaxF: 0.863 -> 0.868Improved

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