NICO++: Towards Better Benchmarking for Domain Generalization
This work addresses the need for better benchmarking in domain generalization, which is crucial for improving the robustness of deep neural networks under distribution shifts, though it is incremental as it builds on existing datasets and methods.
The authors tackled the problem of evaluating domain generalization algorithms by proposing NICO++, a large-scale benchmark with extensive labeled domains and more rational evaluation methods, which demonstrated superior evaluation capability and alleviated unfairness in model selection.
Despite the remarkable performance that modern deep neural networks have achieved on independent and identically distributed (I.I.D.) data, they can crash under distribution shifts. Most current evaluation methods for domain generalization (DG) adopt the leave-one-out strategy as a compromise on the limited number of domains. We propose a large-scale benchmark with extensive labeled domains named NICO++ along with more rational evaluation methods for comprehensively evaluating DG algorithms. To evaluate DG datasets, we propose two metrics to quantify covariate shift and concept shift, respectively. Two novel generalization bounds from the perspective of data construction are proposed to prove that limited concept shift and significant covariate shift favor the evaluation capability for generalization. Through extensive experiments, NICO++ shows its superior evaluation capability compared with current DG datasets and its contribution in alleviating unfairness caused by the leak of oracle knowledge in model selection.