用级联注意力结构提升场景文字识别效率，算力减半精度不变
Efficient and Accurate Scene Text Recognition with Cascaded-Transformers
- 采用分层压缩视觉令牌的级联变压器架构，逐步减少冗余信息
- 大模型下准确率92.68%（原92.77%），计算量几乎减半
- 适合移动端或边缘设备部署的高效高精度文字识别任务
近年来，结合文本解码器的视觉变换器在场景文字识别（STR）中表现出色，因其能捕捉长程依赖与上下文关系，具备高学习能力。然而，这类模型的计算与内存开销巨大，限制了其在资源受限场景的应用。为此，我们提出一种高效且准确的STR系统。核心在于改进编码器结构，引入级联变压器设计：在编码过程中逐层缩减视觉令牌数量，有效去除冗余信息，降低计算成本。实验表明，该系统在保持与当前最优基线相当性能的同时，显著降低计算需求。尤其对于大模型，准确率从92.77%降至92.68%，但计算复杂度几乎减半。
原文摘要 · Abstract (English)
In recent years, vision transformers with text decoder have demonstrated remarkable performance on Scene Text Recognition (STR) due to their ability to capture long-range dependencies and contextual relationships with high learning capacity. However, the computational and memory demands of these models are significant, limiting their deployment in resource-constrained applications. To address this challenge, we propose an efficient and accurate STR system. Specifically, we focus on improving the efficiency of encoder models by introducing a cascaded-transformers structure. This structure progressively reduces the vision token size during the encoding step, effectively eliminating redundant tokens and reducing computational cost. Our experimental results confirm that our STR system achieves comparable performance to state-of-the-art baselines while substantially decreasing computational requirements. In particular, for large-models, the accuracy remains same, 92.77 to 92.68, while computational complexity is almost halved with our structure.
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。