Research map / Continual learning / Domain and prompt-based
A Memorizing and Generalizing Framework for Lifelong Person Re-Identification
Nan Pu, Zhun Zhong, Nicu Sebe, Michael S. Lew
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add accumulated knowledge graph (AKA) with stability-plasticity loss to LwF baseline (ablation) | LwF baseline · Alpha-LReID Order-1 imbalanced | unseen-domain generalization | Improved |
| add plasticity loss Lp to the GCN baseline (ablation) | Baseline + GCN · Alpha-LReID Order-1 imbalanced | unseen-domain accuracy | Improved |
| add stability loss Ls on top of Lp (ablation) | AKA · Alpha-LReID Order-1 imbalanced | seen and unseen accuracy | Improved |
| add fixed-weight ranking distillation Lsfd (ablation) | baseline · Alpha-LReID Order-1 imbalanced | seen vs unseen accuracy | Mixed |
| add dynamic-weight ranking consistency distillation Lrcd (ablation) | baseline · Alpha-LReID Order-1 imbalanced | seen and unseen accuracy | Improved |
| combine AKA with RCD (ablation) | MEGE · Alpha-LReID Order-1 imbalanced | seen and unseen accuracy | Improved |
| detach gradient of the graph network from the feature extractor (ablation) | AKA · Alpha-LReID Order-1 imbalanced | all metrics | Improved |
| stack more GCN layers (ablation) | AKA · Alpha-LReID Order-1 imbalanced | anti-forgetting accuracy | No clear effect |
| learn position weights adaptively instead of fixed weighting strategies (ablation) | RCD · Alpha-LReID Order-1 imbalanced | accuracy | Improved |
| re-initialize position weights at each stage instead of inheriting (ablation) | RCD · Alpha-LReID Order-1 imbalanced | accuracy | Improved |
| use parameterized edge weights for ISG and AKG and non-parameterized for knowledge operation (ablation) | AKA · Alpha-LReID Order-1 imbalanced | H-metric | Improved |
| freeze AKG after an early domain instead of continuing to train it (ablation) | MEGE · Alpha-LReID Order-1 imbalanced unseen | unseen accuracy | Hurt |
Closest papers by meaning
- Lifelong Person Re-Identification via Adaptive Knowledge Accumulation (CVPR 2021 · similarity 0.94)
- Lifelong Person Re-identification by Pseudo Task Knowledge Preservation (AAAI 2022 · similarity 0.93)
- Generalising without Forgetting for Lifelong Person Re-Identification (AAAI 2021 · similarity 0.92)
- Distribution-aware Knowledge Prototyping for Non-exemplar Lifelong Person Re-identification (CVPR 2024 · similarity 0.91)
- LSTKC: Long Short-Term Knowledge Consolidation for Lifelong Person Re-identification (AAAI 2024 · similarity 0.91)
- DASK: Distribution Rehearsing via Adaptive Style Kernel Learning for Exemplar-Free Lifelong Person Re-Identification (AAAI 2025 · similarity 0.90)
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.
