WinARDet:以壓縮窗口自回歸實現長時程視覺語言定位
WinARDet: Compressed Window Auto-Regression for Long-Horizon Vision-Language Grounding
WinARDet提出壓縮窗口自回歸機制,透過限制窗口內的自回歸並壓縮過往窗口,減少密集場景中的重複方框輸出。在每張圖像含有 200-2,000 個物件的 RSDense 上,WinARDet 將 RexOmni F1-score(IoU=0.5)由 32.16 提升至 39.36,冗餘率由 35.48% 降至 13.00%。該方法亦適用於平行方框解碼及多邊形生成。
Published on Oct 14
Authors:
,
,
,
,
,
,
,
Abstract
Compressed Window Auto-Regression partitions long answer sequences into windows with restricted autoregression and compressed latent history to prevent attention takeover and box repetition in dense vision-language grounding.
MLLM-based detectors formulate vision-language grounding as autoregressive coordinate generation, yet in dense scenes, they often collapse into a box repetition loop, repeatedly emitting identical boxes until the decoding budget is exhausted. We trace this failure to historical-answer attention takeover: Under full-history autoregression, repeated coordinate tokens in the growing answer history attract disproportionate attention, crowding out the attention sinks needed to attend to new visual evidence. % To break this loop, in this paper, we propose Compressed Window Auto-Regression (WinAR). This new mechanism partitions a long answer sequence into consecutive generation windows and restricts full token-level autoregression to the current window only while compressing each previous window into a fixed set of latent tokens that condition subsequent decoding. This design reduces the repeated exposure to raw historical coordinates that triggers the attention takeover, yet preserves cross-window dependencies essential for coherent multi-object generation. On RSDense, which contains 200-2,000 objects per image, our model WinARDet improves RexOmni F1-score (IoU=0.5) from 32.16 to 39.36 over the baseline, while reducing the redundancy rate from 35.48% to 13.00%. WinARDet also generalizes across decoding strategies and output formats, including parallel box decoding and polygon generation, providing a practical paradigm for long-horizon vision-language grounding across scene densities.
View arXiv page View PDF Project page GitHub Add to collection
Get this paper in your agent:
hf papers read 2610.19419
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 1
Updated 5 days agofacias/WinARDet
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2610.19419 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2610.19419 in a Space README.md to link it from this page.
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co