Deva Ramanan, Neehar Peri, Shu Kong, Anish Madan
We lifted 9 functions out of this paper's own repositories and ran 6 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| anishmadan23/foundational_fsod | canonical | 6 of 9 |
| Function | Status | Where it lives |
|---|---|---|
| get_clip_embeddings | Ran | anishmadan23/foundational_fsod/detic/predictor.py code served (permissive licence) · get_code("1f211df98dc51cee") |
| get_fed_loss_inds_deterministic_with_negs | Ran | anishmadan23/foundational_fsod/detic/modeling/utils.py code served (permissive licence) · get_code("3b2a9d30d2f8d81a") |
| load_class_freq | Ran | anishmadan23/foundational_fsod/detic/modeling/utils.py code served (permissive licence) · get_code("e17a462624fc9dac") |
| match_name_keywords | Ran | anishmadan23/foundational_fsod/detic/custom_solver.py code served (permissive licence) · get_code("c15434906f804d7f") |
| window_partition | Ran | anishmadan23/foundational_fsod/detic/modeling/backbone/swintransformer.py code served (permissive licence) · get_code("f9fd6241d935f07b") |
| window_reverse | Ran | anishmadan23/foundational_fsod/detic/modeling/backbone/swintransformer.py code served (permissive licence) · get_code("fb32094c6dbece71") |
| compute_average_precision | Not yet run | anishmadan23/foundational_fsod/detic/evaluation/oideval.py code served (permissive licence) · get_code("2e0a799ccbd489e7") |
| create_timm_resnet | Not yet run | anishmadan23/foundational_fsod/detic/modeling/backbone/timm.py code served (permissive licence) · get_code("3634f2208a0255d7") |
| get_fed_loss_inds_deterministic2 | Not yet run | anishmadan23/foundational_fsod/detic/modeling/utils.py code served (permissive licence) · get_code("c66a839ca35205e6") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
The era of vision-language models (VLMs) trained on web-scale datasets challenges conventional formulations of "open-world" perception. In this work, we revisit the task of few-shot object detection (FSOD) in the context of recent foundational VLMs. First, we point out that zero-shot predictions from VLMs such as GroundingDINO significantly outperform state-of-the-art few-shot detectors (48 vs. 33 AP) on COCO. Despite their strong zero-shot performance, such foundation models may still be sub-optimal. For example, trucks on the web may be defined differently from trucks for a target application such as autonomous vehicle perception. We argue that the task of few-shot recognition can be reformulated as aligning foundation models to target concepts using a few examples. Interestingly, such examples can be multi-modal, using both text and visual cues, mimicking instructions that are often given to human annotators when defining a target concept of interest. Concretely, we propose Foundational FSOD, a new benchmark protocol that evaluates detectors pre-trained on any external data and fine-tuned on multi-modal (text and visual) K-shot examples per target class. We repurpose nuImages for Foundational FSOD, benchmark several popular open-source VLMs, and provide an empirical analysis of state-of-the-art methods. Lastly, we discuss our recent CVPR 2024 Foundational FSOD competition and share insights from the community. Notably, the winning team significantly outperforms our baseline by 23.3 mAP! Our code and dataset splits are available on GitHub and HuggingFace.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2312.14494")
get_code_for_paper("2312.14494")
have("2312.14494")
Connect an agent — have() is free.