Handling Label Noise via Instance-Level Difficulty Modeling and Dynamic Optimization
This addresses the challenge of label noise in machine learning, which can degrade model generalization, offering a more efficient and scalable solution compared to existing methods.
The paper tackles the problem of deep neural networks degrading under noisy supervision by proposing a two-stage noisy learning framework that uses instance-level difficulty modeling and dynamic optimization, achieving state-of-the-art performance on benchmarks with a nearly 75% reduction in computational time.
Recent studies indicate that deep neural networks degrade in generalization performance under noisy supervision. Existing methods focus on isolating clean subsets or correcting noisy labels, facing limitations such as high computational costs, heavy hyperparameter tuning process, and coarse-grained optimization. To address these challenges, we propose a novel two-stage noisy learning framework that enables instance-level optimization through a dynamically weighted loss function, avoiding hyperparameter tuning. To obtain stable and accurate information about noise modeling, we introduce a simple yet effective metric, termed wrong event, which dynamically models the cleanliness and difficulty of individual samples while maintaining computational costs. Our framework first collects wrong event information and builds a strong base model. Then we perform noise-robust training on the base model, using a probabilistic model to handle the wrong event information of samples. Experiments on five synthetic and real-world LNL benchmarks demonstrate our method surpasses state-of-the-art methods in performance, achieves a nearly 75% reduction in computational time and improves model scalability.