ICCV 2025 accepted papers on Continual learning
32 of the 2,701 papers accepted at ICCV 2025 (1.2%) are on Continual learning, compared with 2.4% at ICCV 2023. They fall into 4 topics.
See ICCV 2025 in the interactive map
Submitting to ICCV? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Class-incremental · 15 papers
- External Knowledge Injection for CLIP-Based Class-Incremental Learning (22 citations)
- Integrating Task-Specific and Universal Adapters for Pre-Trained Model-based Class-Incremental Learning (16 citations)
- Gradient Decomposition and Alignment for Incremental Object Detection (7 citations)
- DuET: Dual Incremental Object Detection via Exemplar-Free Task Arithmetic (3 citations)
- Flexi-FSCIL: Adaptive Knowledge Retention for Breaking the Stability-Plasticity Dilemma in Few-Shot Class-Incremental Learning (3 citations)
- Revisiting Pool-based Prompt Learning for Few-shot Class-incremental Learning (3 citations)
- A Tiny Change, A Giant Leap: Long-Tailed Class-Incremental Learning via Geometric Prototype Alignment (2 citations)
- ESSENTIAL: Episodic and Semantic Memory Integration for Video Class-Incremental Learning (2 citations)
- FG-OrIU: Towards Better Forgetting via Feature-Gradient Orthogonality for Incremental Unlearning (2 citations)
- Learning Yourself: Class-Incremental Semantic Segmentation with Language-Inspired Bootstrapped Disentanglement (2 citations)
- Seeing 3D Through 2D Lenses: 3D Few-Shot Class-Incremental Learning via Cross-Modal Geometric Rectification (2 citations)
- Continual Multiple Instance Learning with Enhanced Localization for Histopathological Whole Slide Image Analysis (1 citation)
and 3 more in the interactive map.
Online and replay-based · 9 papers
- Generalization-Preserved Learning: Closing the Backdoor to Catastrophic Forgetting in Continual Deepfake Detection (13 citations)
- iManip: Skill-Incremental Learning for Robotic Manipulation (7 citations)
- Joint Diffusion Models in Continual Learning (7 citations)
- CODE-CL: Conceptor-Based Gradient Projection for Deep Continual Learning (3 citations)
- Continual Personalization for Diffusion Models (3 citations)
- Task-Aware Prompt Gradient Projection for Parameter-Efficient Tuning Federated Class-Incremental Learning (2 citations)
- CRAM: Large Scale Video Continual Learning with Bootstrapped Compression (1 citation)
- FedAGC: Federated Continual Learning with Asymmetric Gradient Correction (1 citation)
- PROL : Rehearsal Free Continual Learning in Streaming Data via Prompt Online Learning (1 citation)
Continual fine-tuning of LLMs · 7 papers
- SMoLoRA: Exploring and Defying Dual Catastrophic Forgetting in Continual Visual Instruction Tuning (23 citations)
- Mind the Gap: Preserving and Compensating for the Modality Gap in CLIP-Based Continual Learning (22 citations)
- Achieving More with Less: Additive Prompt Tuning for Rehearsal-Free Class-Incremental Learning (6 citations)
- PLAN: Proactive Low-Rank Allocation for Continual Learning (6 citations)
- Dynamic Multi-Layer Null Space Projection for Vision-Language Continual Learning (4 citations)
- Lark: Low-Rank Updates After Knowledge Localization for Few-shot Class-Incremental Learning
- Any-SSR: How Recursive Least Squares Works in Continual Learning of Large Language Model
