发表机构
Bangladesh Army University of Science and Technology（孟加拉国陆军科技大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文构建专家验证的RSBdSL38数据集，提出轻量型注意力卷积网络，实现高准确率孟加拉语手语识别，模型可在普通智能手机部署，相关资源已公开。
AI 中文摘要
孟加拉国的聋人和听障人士主要通过孟加拉语手语（BdSL）进行交流，在个人设备上实现自动BdSL识别可拓宽其获取教育与服务的渠道。现有系统使用的数据集来自受控场景，且未经过专家验证，同时采用的重型预训练主干网络不适合在设备端使用。本文提出RSBdSL38数据集，包含10874张经专家验证的图像，覆盖全部38种BdSL手势，对应孟加拉语字母表的51个字母，由孟加拉国三所特殊学校的真实手语使用者录制完成。本文提出一种基于注意力的轻量型卷积网络，参数规模为298470，由分组瓶颈残差块、通道与空间注意力模块、多尺度深度手部特征块、双池化层及Swish激活函数构建而成。该模型从零开始训练，准确率达96.37%（5个随机种子的均值为95.72%±0.54%），在相同协议下，与9种基于ImageNet预训练的高效架构中表现最佳的模型相比，准确率仅低1.08个百分点，但参数数量减少8.5至68倍，MACs减少1.3至21.7倍。经重新训练后，该模型在6个公开BdSL基准上的准确率为92.95%至98.33%，在合并语料库上的准确率为97.04%，在BdSL-38上的零样本准确率为76.25%。移除任意一个架构阶段会导致准确率下降7.61至89.30个百分点，而训练配方的影响最多为3.17个百分点。结合删除-插入与权重随机化检验的Grad-CAM分析证实，模型预测依据为手语使用者的手部。在36名手语使用者中留出6名作为独立测试集的情况下，模型准确率为85.18%。将模型量化至0.48 MB后，在普通智能手机上运行时，单张图像推理耗时3.98 ms，内存占用为15.5 MB。综上，RSBdSL38数据集与本文提出的从零开始训练的模型，以远低于预训练主干网络的成本，将基准准确率转化为可部署的易用性；相关数据集、代码与模型已公开。
英文摘要
Deaf and hard-of-hearing people in Bangladesh communicate mainly through Bangla Sign Language (BdSL). Automatic BdSL recognition on personal devices could widen access to education and services. Existing systems use controlled-setting datasets without expert verification and heavyweight pretrained backbones unsuited to on-device use. We introduce RSBdSL38, 10,874 expert-validated images spanning all 38 BdSL hand signs, representing the 51 letters of the Bangla alphabet, recorded from real signers at three special-needs schools across Bangladesh. We propose a lightweight attention based convolutional network of 298,470 parameters, built from grouped bottleneck residual blocks, channel and spatial attention, a multi-scale depthwise hand-feature block, dual pooling, and Swish activations. Trained from scratch, it attains 96.37% accuracy (95.72% +- 0.54% over five seeds), within 1.08 percentage points of the best of nine ImageNet-pretrained efficient architectures under an identical protocol, using 8.5 to 68x fewer parameters and 1.3 to 21.7x fewer MACs. Retrained, it reaches 92.95 to 98.33% on six public BdSL benchmarks, 97.04% on a merged corpus, and 76.25% zero-shot on BdSL-38. Removing any architectural stage costs 7.61 to 89.30 points, against at most 3.17 for the training recipe. Grad-CAM with deletion-insertion and weight-randomization checks confirms that predictions follow the signing hand. A signer-independent split holding out 6 of 36 signers yields 85.18%. Quantized to 0.48 MB, it runs at 3.98 ms per image within a 15.5 MB footprint on a commodity smartphone. Together, RSBdSL38 and our from-scratch model turn benchmark accuracy into deployable accessibility at a fraction of pretrained-backbone cost; dataset, code, and models are released.
