Acoustic levitation system, with the advantages of avoiding clamping force on samples, has been widely introduced into noncontact manipulation. Previous acoustic levitation systems were generally fixed and could not realize remote adjustment of manipulation stage. In this article, we develop an ultrasonic robotic system for small object manipulation which is capable of noncontact operation, remote control, and gesture-based intelligent control. The whole setup is composed of acoustic levitation platform, mobile operation platform, and Kinect sensing system which provides a humanized approach for human-machine interaction. The experimental results validate a sufficient suite of innovations for achieving remote control of this acoustic levitation system based on gesture recognition method. It will not only pave a new way for noncontact manipulation system but also greatly reduce the difficulty and cost of operation.