通过用户反馈动态优化目标检索，提升复杂场景下模糊查询的准确率。
IntRec: Intent-based Retrieval with Contrastive Refinement
- 基于意图状态维护正负样本记忆，实现交互式精炼
- 在LVIS上达35.4 AP， ambiguous任务提升7.9 AP
- 单次反馈后延迟低于30毫秒，适合实时交互应用
从复杂场景中检索用户指定对象仍是挑战性任务，尤其当查询模糊或存在多个相似对象时。现有开放词汇检测器采用一次性推理，无法根据用户反馈进行预测修正。为此，我们提出IntRec，一种基于意图的交互式对象检索框架，通过用户反馈不断优化结果。其核心是意图状态（IS），维护正向锚点（确认线索）与负向约束（被拒绝假设）双重记忆集。对比对齐函数通过最大化与正向线索的相似性、同时惩罚被拒绝项，实现对杂乱场景中物体的细粒度区分。该交互框架在无需额外监督的情况下显著提升检索精度。在LVIS数据集上，IntRec达到35.4 AP，优于OVMR、CoDet和CAKE的+2.3、+3.7、+0.5；在更具挑战性的LVIS-Ambiguous基准上，单次纠正反馈后性能提升7.9 AP，且每次交互延迟不足30毫秒。
原文摘要 · Abstract (English)
Retrieving user-specified objects from complex scenes remains a challenging task, especially when queries are ambiguous or involve multiple similar objects. Existing open-vocabulary detectors operate in a one-shot manner, lacking the ability to refine predictions based on user feedback. To address this, we propose IntRec, an interactive object retrieval framework that refines predictions based on user feedback. At its core is an Intent State (IS) that maintains dual memory sets for positive anchors (confirmed cues) and negative constraints (rejected hypotheses). A contrastive alignment function ranks candidate objects by maximizing similarity to positive cues while penalizing rejected ones, enabling fine-grained disambiguation in cluttered scenes. Our interactive framework provides substantial improvements in retrieval accuracy without additional supervision. On LVIS, IntRec achieves 35.4 AP, outperforming OVMR, CoDet, and CAKE by +2.3, +3.7, and +0.5, respectively. On the challenging LVIS-Ambiguous benchmark, it improves performance by +7.9 AP over its one-shot baseline after a single corrective feedback, with less than 30 ms of added latency per interaction.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。