OODBench: Out-of-Distribution Benchmark for Large Vision-Language Models
This addresses the problem of safety risks in real-world applications like autonomous driving or medical assistance by providing a benchmark for OOD evaluation, though it is incremental as it builds on existing VLM research.
The authors tackled the lack of benchmarks for evaluating large vision-language models on out-of-distribution (OOD) data by proposing OODBench, a method that includes 40K instance-category pairs and an automated assessment metric, showing that current models exhibit notable performance degradation on this benchmark.
Existing Visual-Language Models (VLMs) have achieved significant progress by being trained on massive-scale datasets, typically under the assumption that data are independent and identically distributed (IID). However, in real-world scenarios, it is often impractical to expect that all data processed by an AI system satisfy this assumption. Furthermore, failure to appropriately handle out-of-distribution (OOD) objects may introduce safety risks in real-world applications (e.g., autonomous driving or medical assistance). Unfortunately, current research has not yet provided valid benchmarks that can comprehensively assess the performance of VLMs in response to OOD data. Therefore, we propose OODBench, a predominantly automated method with minimal human verification, for constructing new benchmarks and evaluating the ability of VLMs to process OOD data. OODBench contains 40K instance-level OOD instance-category pairs, and we show that current VLMs still exhibit notable performance degradation on OODBench, even when the underlying image categories are common. In addition, we propose a reliable automated assessment metric that employs a Basic-to-Advanced Progression of prompted questions to assess the impact of OOD data on questions of varying difficulty more fully. Lastly, we summarize substantial findings and insights to facilitate future research in the acquisition and evaluation of OOD data.