FlexiGrad: Adaptive Gradient Modulation for Hierarchical Fine-Grained Classification

arXiv:2607.17563v2 Announce Type: replace Abstract: Many fine-grained recognition tasks contain hierarchical labels such as order, family and species. Although this supervision should be beneficial, jointly optimising all levels often leads to unstable training because coarse and fine classifiers impose inconsistent gradients on the shared backbone. This hierarchical gradient conflict prevents the model from learning a coherent coarse-to-fine representation. In this paper, we propose FlexiGrad, a simple and parameter-free method that regulates gradient interactions during backpropagation. FlexiGrad removes only the harmful conflicting component when tasks disagree and reinforces the shared direction when they partially agree through a smooth hierarchy-aware weighting function. This produces stable optimisation and preserves both global structure and fine-grained discriminative cues. FlexiGrad integrates into existing architectures without modification while improves multi-granularity accuracy on CUB-200-2011, FGVC-Aircraft and Stanford Cars. The code will be available at PRIS-CV/FlexiGrad.

Related

• Free-Grained Hierarchical Visual Recognition
• How many labels do you need? A decision framework for cross-habitat marine species recognition
• SARE: Sample-wise Adaptive Reasoning for Training-free Fine-grained Visual Recognition

Source: arXiv cs.CV | 2026-07-31