利用 AI 革新零售业

减少损耗,改善预测,推动收入增长

领先的零售商正利用 AI 来减少损耗、改善预测、实现仓库物流自动化、决定店内促销活动和实时定价、为客户提供个性化服务和推荐,以及在实体店和网店提供更出色的购物体验。

  • <span style=

    智能商店

  • <span style=

    需求预测

  • <span style=

    仓库物流

  • <span style=

    推荐系统和搜索

  • <span style=

    会话式 AI

会话式 AI

Intelligent Stores

通过使用来自摄像头和传感器的数据,零售商正利用 AI 来减少损耗、消除缺货问题,以及清楚查看顾客行为。同样的基础设施也能加快结账速度。研究零售商如何使用 AI 从以下四个方面打造智能商店:资产保护、商店分析、自助购物和商店运营。

Forecasting and Inventory Management

预测和库存管理

AI 还能改进需求预测和库存管理。需求预测使用来自各种来源的数据,以确保在合适的时间和合适的商店发售合适的产品。​

使用机器学习提高预测准确度能对供应链优化产生重大影响。

有效的预测不仅仅要考虑人口统计数据和地点。许多外部事件(例如天气或当地体育赛事)也可能会影响供给和需求。通过基于 NVIDIA GPU 运行 NVIDIA RAPIDS™ 软件库,零售商可以将机器学习算法的训练速度提升高达 20 倍。这意味着他们可以使用更多的数据,并以更高的准确度更快地处理数据。​​

沃尔玛改善预测

沃尔玛改善预测

以沃尔玛为例,借助 RAPIDS 开源数据处理和机器学习库,该公司已将机器学习算法的训练速度提升 20 倍。RAPIDS 基于 CUDA-X AI™ 并利用 NVIDIA GPU,借此沃尔玛能够更加高效地将正确的产品送到正确的商店,实时响应购物者趋势,并大规模节省库存成本。

了解沃尔玛如何改善预测水平
仓库物流领域的 AI

仓库物流领域的 AI

借助深度学习和增强学习,可以在供应链中将 AI 与仓库管理结合起来。这种结合为运营增加了一定程度的“意识”,让仓库机器人能够从整体角度看待产品(同时利用物理特性和理论特性),以做出明智的决策。​

NVIDIA EGX™ 平台通过在边缘位置(位于仓库内)部署 GPU 加速的 AI,提高仓库生产力、效率和订单准确性。负责处理订单的仓库机器人由 GPU 驱动,不仅能实时运行,还能在做出决策之前测量所有变量和就地适应变化的情况。它们能够自动生成报告(可能包括需求预测和最后一公里交付),从而提供端到端可见性,并提高按订单拣货、包装和发运的准确性。

推荐和视觉搜索

推荐和视觉搜索

 了解顾客行为对于希望推动业务发展的零售商而言变得更为重要。利用视频分析技术驱动的 AI 应用程序,零售商可以清楚看到店内的顾客行为,就像他们当前在线看到的一样。​

通过深入了解客流量大的过道、停留时间和人口统计数据,零售商可以改善推销效果和在店内举行实时促销活动,以增加收入并提供更好的体验。

对于电子商务,零售商正使用 GPU 驱动的机器学习和深度学习算法来打造更快、更准确的推荐引擎,从而将收入提升 60%。

会话式 AI

会话式 AI


自然语言处理 (NLP) 正帮助零售商为顾客提供个性化体验及改善顾客服务。它还用于汇编和分析消费者数据,以获得切实可行的见解。NVIDIA 可优化 BERT(一种热门的 NLP 模型)的训练和推理性能,从而实现实时会话式 AI。

NVIDIA Jarvis 是一种用于构建和部署 AI 应用程序的平台,能够融合用于语音识别和语音合成、语言理解和视觉的深度学习模型。Jarvis 在 NVIDIA EGX 堆栈上运行,该堆栈与所有商用 Kubernetes 基础设施兼容。

Do Your Life’s Work from Anywhere

As the world shifts to remote work, consumer buying trends are shifting with it—and retail professionals from IT managers to data scientists need to keep up with those trends. Explore how GPU-powered hardware and virtualization software can help your organization digitally transform and keep pace with a changing world.

Read E-Book: Do Your Life’s Work from Anywhere in Retail

View Webpage: Work Remote with NVIDIA

Trailblazing Retail Startups

These innovators are the powerful engineers of several new creations benefitting the retail industry. As part of the NVIDIA Inception program, NVIDIA’s startup incubator, they’ve developed game-changing, GPU-based AI tools for retail.  Inception nurtures the entrepreneurship ecosystem by connecting pioneers to a massive network of deep learning experts and thought leaders. Explore some of the most fashionable applications at the forefront of the fourth Industrial Revolution. 

AnyVision—Streamlining Access with Intelligent Video Analytics

AnyVision – 通过面部识别简化访问过程

AnyVision 正使用 AI 生成的热图进行智能店内状况分析,以改善用户体验并优化零售商店的内部布局。

Malong Technologies—Classifying Unlabeled Images

码隆科技公司 – 将无标记的图像分类

码隆科技公司正在提高效率,并使用智能视频分析技术准确减少店内损耗和改进自助服务。

NVIDIA AI Consulting Partner Network

Service Delivery Partners have special expertise in the transformational business benefits afforded by deep learning (DL), machine learning (ML), and artificial intelligence (AI). Learn about the consulting services they provide, as well as impactful solutions specifically for critical use cases in retail like asset protection and predictive analytics.

AI Retail from Edge to Data Center to Cloud

Real-Time AI at the Edge

Real-Time AI at the Edge

NVIDIA’s edge solutions are designed to gather and compute continuous streams of data at the network’s edge. AI computations are performed entirely in the store, delivering real-time insights and notifications to store associates on shrinkage and providing insights into customer demographics, shopping preferences, and more.

Smart retail is possible with today’s powerful AI and the NVIDIA EGX platform, which brings the power of accelerated AI computing to retail stores.

Powerful Computing for Data Centers

Powerful Computing for Data Centers

NVIDIA Tensor Core GPUs dramatically speed up deep learning and machine learning model training to deliver never-before-possible insights. From edge to data center, Tensor Core GPUs are available from every major computer system and server manufacturer to accelerate training of AI models. They’re also available in NVIDIA DGX™ systems, which are equipped with the DGX software stack for rapid AI deployment to meet the demands of deep learning and machine learning developers.

NVIDIA DGX™ A100 is the universal system for all AI workloads, offering unprecedented compute density, performance, and flexibility in the world’s first 5 petaFLOPS AI system.

Democratization, from Data Center to the Cloud

Democratization, from Data Center to the Cloud

NVIDIA GPUs are available in all major cloud platforms worldwide, and NGC provides GPU-accelerated software containers for easy deployment, including deep learning frameworks like TensorFlow, PyTorch, MXNet, and more. NVIDIA Metropolis is also available in the cloud, fully-integrated with Azure IoT Edge and is soon-to-be integrated with AWS IoT Greengrass.

NVIDIA’s software libraries and SDKs create a scalable solution that enables customers to deploy inference and AI in the cloud, on their servers, or at the edge. These SDKs include NVIDIA JetPack for embedded, DeepStream for IVA, NVIDIA Isaac for robotics, NVIDIA TensorRT for inference, Transfer Learning Toolkit for tuning deep neural networks (DNNs), and NGC for containers and AI software.

Sign up for the latest retail news from NVIDIA.