Local control of sound around human ears in complex acoustic environments is important for both active noise control and sound reproduction. Two typical active noise control approaches for this objective are active headrest systems and virtual sound barrier systems. In this paper, the history and the present status for the active headrest systems and virtual sound barrier systems are briefed first, then the theoretical principles, the design methods and the applications of these approaches are reviewed. Their advantages and limitations are discussed, and finally, the currently existing problems and future research directions are presented. The feasibility of these approaches to generating a quiet zone near a human ear has been verified by the theoretical research, numerical simulations and experiments. The active headrest systems require less control sources and are simpler for implementation; however, they suffer the problem of small-sized quiet zones. This results in the restrictions on the head movement since the error sensor needs to be close to the human ear to obtain better noise reduction performance. Based on the virtual sensor technology, a physical error sensor can be placed farther away from the human head, and create the quiet zone at the virtual sensor position near the human ear. Moreover, combined with the virtual sensor technology and the head-tracking technology, an active headrest system can generate a moving zone of quiet following the head movement, and the noise reduction can be achieved in a middle-to-high frequency range. A virtual sound barrier system reduces the sound pressure inside a volume, through controlling the sound pressure and normal gradient on the boundary of the volume. Two main design methods are the expansion method of the primary sound field which is suitable for steady primary sound fields, and the least mean square method which is applicable to time-varying primary sound fields. It can generate larger quiet zone at the cost of more control sources, more complexity and high cost. Optimizing cost functions and control sources and using hybrid active and passive control techniques can increase the effective frequency range and reduce the number of control sources. Although the feasibility of these two systems has been verified, more research work is needed to develop practical systems. An active-passive hybrid structure for specific application scenarios, which combines these two approaches together as well as the virtual sensor technology and sound field estimation technology, may most likely be practical methods to achieve effective noise reduction near the human ear in a complex sound field in the near future.