arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2212.03474cs.LGcs.AI
收藏

Tree DNN:一种深度容器网络

Tree DNN: A Deep Container Network

Sony Research India(索尼印度研究院);JIIT(贾伊佩信息技术学院);Microsoft Research, India(微软印度研究院);Samsung Research Institute, Bangalore(三星班加罗尔研究院)

Brijraj Singh, Swati Gupta, Mayukh Das, Praveen Doreswamy Naidu, Sharan Kumar Allur

摘要:

多任务学习(MTL)已经在用户产品中显示出其重要性,以实现快速训练,数据效率,减少过拟合等。MTL通过共享网络参数并同时为多个任务训练网络来实现这一点。然而,MTL并不提供解决方案,如果每个任务都需要从不同的数据集进行训练。为了解决上述问题,我们提出了一种名为TreeDNN的架构及其训练方法。TreeDNN有助于同时使用多个数据集训练模型,其中树的每个分支可能需要不同的训练数据集。我们在结果中表明,TreeDNN提供了有竞争力的性能,其优点是减少了对参数存储的ROM要求,并通过在推理时仅加载特定分支来提高系统的响应性。

英文摘要:

Multi-Task Learning (MTL) has shown its importance at user products for fast training, data efficiency, reduced overfitting etc. MTL achieves it by sharing the network parameters and training a network for multiple tasks simultaneously. However, MTL does not provide the solution, if each task needs training from a different dataset. In order to solve the stated problem, we have proposed an architecture named TreeDNN along with it's training methodology. TreeDNN helps in training the model with multiple datasets simultaneously, where each branch of the tree may need a different training dataset. We have shown in the results that TreeDNN provides competitive performance with the advantage of reduced ROM requirement for parameter storage and increased responsiveness of the system by loading only specific branch at inference time.

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。

↑