AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

OpenSRH: optimizing brain tumor surgery using intraoperative stimulated Raman histology

Cheng Jiang, Asadur Zaman Chowdury, Xinhai Hou, Akhil Kondepudi, Christian Freudiger, Kyle Stephen Conway, Sandra Camelo-Piragua, Daniel A Orringer, Honglak Lee, Todd Hollon

NeurIPS 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 12 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 ImageNet pretraining instead of random initResNet50 · OpenSRH patch-level valaccuracy: 84.4 -> 86.5Improved
use ImageNet pretraining instead of random initViT-S · OpenSRH patch-level valaccuracy: 77.2 -> 83.7Improved
use ImageNet pretraining instead of random initResNet50 · OpenSRH patient-level valaccuracy: 90.0 -> 88.9Hurt
use ImageNet pretraining instead of random initViT-S · OpenSRH patient-level valaccuracy: 85.0 -> 88.9Improved
use ResNet50 instead of ViT-Srandom init · OpenSRH patch-level valaccuracy: 77.2 -> 84.4Improved
SimCLR pretraining on OpenSRH instead of ImageNet transferResNet50 linear eval · OpenSRH patch-level valaccuracy: 68.3 -> 79.1Improved
SimCLR pretraining on OpenSRH instead of ImageNet transferViT-S linear eval · OpenSRH patch-level valaccuracy: 71.8 -> 76.8Improved
SupCon instead of SimCLRResNet50 linear eval · OpenSRH patient-level valaccuracy: 83.9 -> 90.0Improved
SupCon instead of SimCLRViT-S linear eval · OpenSRH patient-level valaccuracy: 80.0 -> 87.8Improved
SupCon instead of SimCLRViT-S fine-tuning · OpenSRH patch-level valaccuracy: 81.4 -> 81.2No clear effect
SupCon instead of ImageNet pretrainingViT-S linear eval · OpenSRH patient-level valaccuracy: 88.3 -> 87.8No clear effect
fine-tune backbone instead of linear eval (SimCLR)ResNet50 · OpenSRH patch-level valaccuracy: 79.1 -> 86.3Improved

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