Beyond Model Size: Probing the Gaps in Visual in-Context Learning by Training a Tiny Model
About
Visual in-Context Learning (VICL) aims at making progress towards adaptive vision models, that can -- based on a few examples -- adapt to a new task at test-time. With the history of in-context learning in natural language processing research, where large, parameter-heavy models are in use, one pathway that current VICL methods take is model- and data-scaling as key ingredients. Yet, it is not clear, whether these ingredients are the key for in-context learning to take shape in vision models. To stress-test such large models, we challenge them with an extreme counterexample: we train a tiny visual in-context model with merely $1$ million parameters and a modest amount of $70,000$ images. We compare the results of this severely capacity capped tiny model to $7,000\times$ larger VICL models in different adaptive settings, (1) on image data with small distribution shifts, (2) on unseen task encodings and (3) on a completely new task, i.e., the setting VICL envisions. With the chasm of training resources between the tiny- and large models, our experiments showcase a lack in how adaptive capabilities are measured, with respect to how tasks are encoded, which tasks were used in pre-training and the choice of metrics. These gaps in current VICL benchmarking underscore a need for innovation in evaluation of adaptive capabilities.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Depth Estimation | NYU Depth V2 | RMSE0.5526 | 226 | |
| Edge Detection | BSDS500 | Mean Precision0.2587 | 15 | |
| De-raining | DID-MDN | LPIPS0.1247 | 7 | |
| Image Colorization | ImageNet | LPIPS0.2672 | 7 | |
| Semantic segmentation | Pascal VOC (split 0) | mIoU13.19 | 7 | |
| Semantic segmentation | PASCAL VOC Split 1 | mIoU16.69 | 7 | |
| Semantic segmentation | PASCAL VOC Split 3 | mIoU13.5 | 7 | |
| Semantic segmentation | PASCAL VOC Split 2 | mIoU12.57 | 7 |