AI Paper Atlas

Research map / Continual learning / Class-incremental

Incremental Learning of Object Detectors Without Catastrophic Forgetting

Konstantin Shmelkov, Cordelia Schmid, Karteek Alahari

ICCV 2017 · Continual learning › Class-incremental · approach: Incremental object detection with replay · 575 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 L2 distillation loss on old-class logits (ablation)B(20) incremental detector · VOC 2007 test (add 1 class tvmonitor)mAP old classes: 25.0 -> 68.3Improved
add L2 distillation loss on old-class logits (ablation)B(11-20) incremental detector · VOC 2007 test (add 10 classes)mAP old classes: 12.8 -> 63.2Improved
finetune the conv trunk instead of freezing it (ablation)B(20) with distillation · VOC 2007 test (add 1 class tvmonitor)mAP new class: 43.2 -> 58.3Improved
add bounding box regression distillation (ablation)B(11-20) · VOC 2007 test (add 10 classes)mAP all classes: 60.9 -> 63.1Improved
add bounding box regression distillation (ablation)B(20) · VOC 2007 test (add 1 class tvmonitor)mAP new class: 62.7 -> 58.3Hurt
use L2 distillation instead of cross-entropy distillation (ablation)B(20) · VOC 2007 test (add 1 class tvmonitor)mAP new class: 52.0 -> 58.3Improved
use L2 distillation instead of cross-entropy distillation (ablation)B(20) · VOC 2007 test (add 1 class tvmonitor)mAP old classes: 67.3 -> 67.8No clear effect
sample non-background (lowest background score) proposals instead of random ones for distillation (ablation)B(16)...(20) sequential · VOC 2007 test (add 5 classes)mAP all classes: 46.0 -> 62.4Improved
add all new classes at once instead of sequentially (ablation)B(16-20) · VOC 2007 test (add 5 classes)mAP all classes: 62.4 -> 65.9Improved
use distillation instead of EWC (ablation)B(11-20) · VOC 2007 test (add 10 classes)mAP all classes: 46.3 -> 63.1Improved
train second stage on COCO images instead of VOC images (ablation)B(11-20) · VOC 2007 test (add 10 classes, second stage data)mAP all classes: 63.2 -> 54.8Hurt
train one separate network per new class instead of one incremental network (ablation)A(1-15)+5 classes · VOC 2007 test (add 5 classes)mAP all classes: 65.9 -> 62.4Hurt

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