Search papers, labs, and topics across Lattice.
This paper introduces contrastive order learning (ConOrd), a novel framework that combines the benefits of contrastive learning and order learning to enhance ordinal regression tasks. By employing a contrastive order loss with soft affinity and disparity weights, ConOrd effectively captures global ordinal structures while leveraging all samples in a batch, overcoming the limitations of traditional methods. Extensive experiments across various tasks, including facial age estimation and quality assessments, show that ConOrd achieves state-of-the-art performance and demonstrates strong generalization capabilities.
ConOrd redefines ordinal regression by seamlessly integrating global structure into contrastive learning, outperforming existing methods across multiple benchmarks.
We propose contrastive order learning (ConOrd), a contrastive learning framework for ordinal regression that integrates the strengths of contrastive learning and order learning. While contrastive learning effectively leverages all samples in a batch, it typically ignores the inherent ordering among rank labels. Conversely, order learning explicitly models label ordinality but often relies on local, margin-based comparisons, limiting its ability to capture global ordinal structure. ConOrd addresses these limitations by introducing a contrastive order loss with soft affinity and disparity weights based on rank differences, enabling fine-grained modeling of ordinal relationships across all sample pairs within a batch. Extensive experiments on a range of ordinal regression tasks, including facial age estimation, blind image quality assessment, and blind video quality assessment, demonstrate that ConOrd consistently achieves state-of-the-art performance and generalizes well across diverse ordinal regression scenarios. The source code is available at https://github.com/cwlee00/ConOrd.
