Customer Service Automation
Deep Robotics needed dependable long-distance voice interaction for its quadruped robots, but far-field use in noisy, reverberant 2-5 meter environments made wake-word detection and instruction accuracy difficult.
Reported impact
Deep Robotics improved the realism and reliability of robot interaction, gained stronger wake-word and localization performance in noisy environments, and reduced the amount of internal effort required for acoustic optimization.
Deep Robotics used Baidu AI Cloud to harden the voice-interaction stack inside the Jueying Lite3 robot. The implementation pairs Baidu's Qianfan model-development and speech stack with an end-to-end speech-language model optimized for intelligent hardware. The voice layer is tuned for far-field human-robot interaction, where low signal-to-noise ratios, moving microphones and variable robot orientation would otherwise break recognition quality. Baidu supplied acoustic-structure optimization, wake-word customization and cloud content integration so Lite3 could support voice wake-up, natural dialogue, body-following scenarios and consumer-style content requests in a single robot experience. In operation, user speech is captured by the robot's microphone array, processed through Baidu's speech and localization pipeline, then passed to the robot interaction logic and connected cloud content services. This let Deep Robotics focus on product behavior and service design while outsourcing the hardest parts of the speech front end to a stack already proven in deployed devices.
Voice commands are captured by the robot in 2-5 meter far-field conditions, run through Baidu's speech and localization pipeline for wake-up and recognition, then passed to robot interaction logic and cloud content services so Lite3 can answer, react and guide follow-on behaviors.
Baidu AI Cloud provided a far-field human-robot voice-interaction solution for the Jueying Lite3 series (original: 百度智能云为云深处科技绝影Lite3系列供远场人机语音交互方案).
Baidu's acoustic-structure design optimization lets the robotics team focus more on product-function definition (original: 百度提供的声学结构设计优化技术支持，让云深处机器人产品享有业界领先的声学效果，云深处科技研发团队可更聚焦于产品功能定义).
In the Jueying Lite3 voice-interaction scenario, the typical distance is 2 to 5 meters, creating low signal-to-noise, background-noise and reverberation challenges (original: 在与绝影Lite3机器人的语音交互场景中，通常是2~5米的远场场景，这带来了一系列的复杂挑战).
Under high background noise, wake-up recognition exceeds 90% and 5-meter localization reaches ±15 degrees (original: 在高背噪工况下，语音唤醒识别率高于90%，并具备360度远场声源定位，在5米范围内可实现±15度高精度定位).
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
In-house Sample Aug 27, 2026 Lenovo Lenovo needed to handle customer support at very large scale across its global device support operation, where routine issues created cost and complexity and human-led resolutions costing $30 to $50 could delay simple… $30 to $50 Lenovo Intelligent Voice Agent · Lenovo Machines Fix Machines Agent · Lenovo Smart Repair Agent 
Atos Sample Jul 29, 2026 Misumi USA Misumi USA needed to handle about 2,000 inbound emails and more than 3,000 attached documents each day, but manual routing and document handling caused frequent misrouting, slowed order and quote processing, and made it… 14 years reduced to 2 years for web category transformation AWS Lambda / API gateway · Amazon Bedrock · Amazon Connect 
ServiceNow Sample Archived · Jul 21, 2026 Lenovo Lenovo's Cloud Service Provider managed-services desk was handling 95% of cases across six mailboxes, with customer data scattered across multiple systems and SharePoint, making software-support cases slow to open, hard… 1 Platform instead of six mailboxes to track customer activities Now Assist for TPSM · ServiceNow AI Platform · Technology Provider Service Management 
