Sensing and localization mechanisms, employed by mobile robots for the detection of obstacles and other nearby agents, may inaccurately estimate the position of obstacles due to noise, delays, and interferences incurred during the detection process. Therefore, it is critical to design collision avoidance strategies that are robust to the presence of measurement errors. In this paper, we present a decentralized, cooperative collision avoidance strategy for a pair of agents considering bounded sensing uncertainties and acceleration constraints. The avoidance control can be appended to any other stable control law (i.e., main control objective) and is active only when the vehicle is close to the other agent. A numerical example is presented that validates the proposed avoidance strategy.