CORTEXA
← Browse
arxivcs.CV2026-07-02

DRDN: Decoupled Representation Dynamic Network for From-Scratch ViT Class-Incremental Learning

Bingchen Huang, Yifu Chen, Zhiling Wang, Yuanchao Du

Dynamic expansion methods for class-incremental learning (CIL) protect task-specific knowledge by growing dedicated tokens or subnetworks, yet our analyses suggest that classification supervision alone does not sufficiently preserve task-agnostic shared backbone representations over long incremental sequences. We identify two intertwined challenges: cross-task confusion from sequential training on predominantly current-task data, which biases decision boundaries toward recent tasks; and under-optimized shared representations in the backbone that cap long-term discriminability as tasks accumulate. We propose the Decoupled Representation Dynamic Network (DRDN), which addresses these challenges via two orthogonal mechanisms. For shared backbone representations, DRDN continuously applies masked image modeling (MIM) at every incremental step, with reconstruction gradients routed exclusively through the backbone, encouraging it to retain general visual structure beyond class-discriminative cues. For task-specific discrimination, DRDN employs hierarchical task token expansion across all transformer layers, with a modified per-task attention rule that reduces inter-task interference. We support this design with accuracy degradation analysis and cross-task confusion rate measurements. In the from-scratch ViT CIL setting (no external pretraining), DRDN consistently improves over strong token-expansion baselines with comparable backbone scale. On CIFAR100-B0 (10 steps), DRDN achieves 77.19% average accuracy, outperforming DKT by 1.36 points and DyTox by 3.53 points, with an advantage that grows at longer incremental sequences. Multi-seed validation confirms stability (+/-0.31%). The MIM decoder is active only during training, adding no inference-time parameters or computation.

View free PDFSource page

Related papers

arxivcs.MMcs.CV2026-07-24

CARA: Concept-Aware Risk Attention for Interpretable Collision Anticipation

Zhishan Tao, Ruoyu Wang, Yucheng Wu, Enjun Du, Yilei Yuan, Sherwin Ho, et al.

Collision anticipation in autonomous driving requires not only accurate early warnings but also interpretable reasoning about what risk factors are being tracked and how risk evolves over time. Existing methods fall short in this regard: feature-driven models are opaque, post-hoc…

View free PDFSource page
arxivcs.CVcs.RO2026-07-24

JustDepth: Real-Time Radar-Camera Depth Estimation with Single-Scan LiDAR Supervision

Wooyung Yun, Dongwook Kim, Soomok Lee

Accurate yet low-latency depth is essential for radar-camera perception in autonomous systems. Cameras provide rich appearance but lack metric scale, whereas automotive radar offers metric range but is sparse and noisy. Many pipelines are multi-stage or depend on auxiliary annota…

View free PDFSource page