AI 中文总结
研究句子级手语翻译，以实时部署为目标，在How2Sign子集上微调SHuBERT-ByT5翻译堆栈，模型取得一定指标。提出硬件感知流系统，通过多种技术降低响应延迟，捕获协议无关客户端，同一后端可服务多设备。
AI 中文摘要
大多数手语理解系统在孤立手势层面运行，限制了自然交流中的实用性。本文以实时部署为主要目标研究句子级手语翻译，而非提出新翻译架构。在因计算和存储限制而均匀采样的9872个How2Sign示例子集上，使用QLoRA微调SHuBERT-ByT5翻译堆栈，同时冻结SHuBERT。模型验证BLEU为16.7，测试集BLEU为15.9，BLEURT为44.7。主要贡献是硬件感知流系统，Raspberry Pi 4B参考客户端提供相机捕获、本地文本显示和语音输出，计算密集型感知和翻译在CPU/GPU后端运行。捕获协议与客户端无关，相同后端可服务多种设备。通过多种技术，完整9872示例工作子集的最终响应平均延迟从1.873秒降至1.354秒（降低27.71%），P95延迟从2.919秒降至2.130秒（降低27.03%）。
英文摘要
Most sign language understanding systems operate at the level of isolated signs, limiting their usefulness in natural communication. We study sentence-level sign language translation (SLT) with the primary goal of real-time deployment rather than proposing a new translation architecture. We fine-tune a SHuBERT-ByT5 translation stack on a uniformly sampled 9,872-example subset of How2Sign, selected because of compute and storage constraints, using QLoRA while keeping SHuBERT frozen. The model obtains a validation BLEU of 16.7 and, on the test split, BLEU 15.9 and BLEURT 44.7. The main contribution is a hardware-aware streaming system: a Raspberry Pi 4B reference client provides camera capture, local text display, and speech output, while compute-intensive perception and translation run on a CPU/GPU backend. The capture protocol remains client-agnostic, so the same backend can serve a browser, phone, or laptop. Chunked ingestion, bounded queues, parallelized perception, temporal reordering, and a sentence-boundary state machine reduce mean post-finalization response latency from 1.873 to 1.354 seconds (27.71%) and P95 latency from 2.919 to 2.130 seconds (27.03%) over the complete 9,872-example working subset.
Comments8 pages, 4 figures, 9 tables
