Research map / Continual learning / Domain and prompt-based
Deep Face Detector Adaptation Without Negative Transfer or Catastrophic Forgetting
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Adapt the source detector using the residual-loss and residual-score method | Faster-RCNN · FDDB | ROC (discrete scores) | Helped |
| Adapt the source detector using the residual-loss and residual-score method | CascadeCNN · FDDB | ROC (discrete scores) | Helped |
| Adapt the source detector using the residual-loss and residual-score method | Faster-RCNN and CascadeCNN · FDDB, semi-supervised adaptation | ROC | Helped |
| Adapt the detector with Gaussian-process score updates | Source face detectors · FDDB, unsupervised adaptation | ROC | No clear effect |
| Add the residual loss (ablation) | Proposed face detector adaptation method · FDDB, supervised adaptation | ROC | Helped |
| Parameterize the target classifier as the source classifier plus an offset (residual score) (ablation) | Proposed face detector adaptation method · FDDB, supervised adaptation | ROC | Helped |
| Fine-tune the detector on target-domain labels without the proposed safeguards | Source face detector · WIDER FACE validation set (Easy, Medium, and Hard) | ROC | Hurt |
| Use residual loss and residual-score adaptation rather than ordinary fine-tuning | Source face detector · WIDER FACE validation set | ROC | Helped |
Closest papers by meaning
- Automatic Adaptation of Object Detectors to New Domains Using Self-Training (CVPR 2019 · similarity 0.88)
- Target-Aware Generative Augmentations for Single-Shot Adaptation (ICML 2023 · similarity 0.86)
- Diversify and Match: A Domain Adaptive Representation Learning Paradigm for Object Detection (CVPR 2019 · similarity 0.86)
- One-Shot Domain Adaptation for Face Generation (CVPR 2020 · similarity 0.85)
- SCAN: Cross Domain Object Detection with Semantic Conditioned Adaptation (AAAI 2022 · similarity 0.85)
- Link-based Contrastive Learning for One-Shot Unsupervised Domain Adaptation (CVPR 2025 · similarity 0.84)
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.