Pailitao-VL：统一的嵌入与重排序用于实时多模态工业搜索
Pailitao-VL: Unified Embedding and Reranker for Real-Time Multi-Modal Industrial Search
AI总结：
Pailitao-VL通过统一嵌入与重排序方法，解决多模态工业搜索中的高精度、实时性和抗噪声问题，实现先进检索架构的高效部署。
摘要：
在本工作中，我们提出了Pailitao-VL，一个全面的多模态检索系统，专为高精度、实时的工业搜索而设计。我们在此针对当前SOTA解决方案中的三个关键挑战：检索粒度不足、易受环境噪声影响以及效率与性能之间的巨大差距。我们的主要贡献在于两个根本性的范式转变。首先，我们将嵌入范式从传统的对比学习转变为绝对ID识别任务。通过将实例锚定到由数十亿语义原型定义的全局一致潜在空间中，我们成功克服了现有嵌入解决方案中固有的随机性和粒度瓶颈。其次，我们将生成性重排序从孤立的点评估转变为比较和校准的列表策略。通过结合基于块的比较推理与校准的绝对相关性评分，系统实现了细腻的辨别分辨率，同时避免了传统重排序方法通常伴随的高延迟问题。广泛的离线基准测试和在线A/B测试表明，Pailitao-VL在阿里巴巴电子商务平台上实现了最先进的性能，并带来了显著的商业影响。这项工作展示了在 demanding、大规模生产环境中部署基于高级 MLLM 的检索架构的稳健且可扩展的路径。
英文摘要：
In this work, we presented Pailitao-VL, a comprehensive multi-modal retrieval system engineered for high-precision, real-time industrial search. We here address three critical challenges in the current SOTA solution: insufficient retrieval granularity, vulnerability to environmental noise, and prohibitive efficiency-performance gap. Our primary contribution lies in two fundamental paradigm shifts. First, we transitioned the embedding paradigm from traditional contrastive learning to an absolute ID-recognition task. Through anchoring instances to a globally consistent latent space defined by billions of semantic prototypes, we successfully overcome the stochasticity and granularity bottlenecks inherent in existing embedding solutions. Second, we evolved the generative reranker from isolated pointwise evaluation to the compare-and-calibrate listwise policy. By synergizing chunk-based comparative reasoning with calibrated absolute relevance scoring, the system achieves nuanced discriminative resolution while circumventing the prohibitive latency typically associated with conventional reranking methods. Extensive offline benchmarks and online A/B tests on Alibaba e-commerce platform confirm that Pailitao-VL achieves state-of-the-art performance and delivers substantial business impact. This work demonstrates a robust and scalable path for deploying advanced MLLM-based retrieval architectures in demanding, large-scale production environments.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
