AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

KD-MVS: Knowledge Distillation Based Self-Supervised Learning for Multi-View Stereo

Yikang Ding, Qingtian Zhu, Xiangyue Liu, Wentao Yuan, Haotian Zhang, Chi Zhang

ECCV 2022 · Label-efficient and robust learning › Knowledge distillation · approach: Few-shot and metric distillation · 40 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 internal featuremetric loss to photometric loss (ablation)teacher model · DTUOverall (mm): 0.495 -> 0.428Improved
use internal instead of external ResNet-18 featuremetric loss (ablation)teacher model · DTUOverall (mm): 0.459 -> 0.428Improved
add external ResNet-18 featuremetric loss to photometric loss (ablation)teacher model · DTUOverall (mm): 0.495 -> 0.459Improved
use second round of distillation (student becomes teacher) (ablation)KD-MVS CasMVSNet · DTUOverall (mm): 0.361 -> 0.327Improved
use more than two distillation rounds (ablation)KD-MVS CasMVSNet · DTUOverall (mm): 0.327 -> 0.328No clear effect
add validated mask to GT-depth training (ablation)student · DTUOverall (mm): 0.352 -> 0.343Improved
replace GT depth with validated pseudo depth under l1 loss (ablation)student · DTUOverall (mm): 0.343 -> 0.346No clear effect
replace l1 loss with probabilistic distillation loss (ablation)student · DTUOverall (mm): 0.346 -> 0.327Improved
increase input views from 3 to 5 (ablation)KD-MVS student · DTUOverall (mm): 0.341 -> 0.327Improved
increase input views from 5 to 9 (ablation)KD-MVS student · DTUOverall (mm): 0.327 -> 0.331No clear effect
raise resolution from 512x640 to 864x1152 (ablation)KD-MVS student · DTUOverall (mm): 0.359 -> 0.327Improved
use stricter cross-view check thresholds (tau_conf 0.30, 2.0, 0.020 to 0.15, 1.0, 0.010) (ablation)KD-MVS student · DTUOverall (mm): 0.396 -> 0.327Improved

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