The reconfigurable intelligent surface (RIS) assisted crowd spectrum detection (CSD) has emerged as a promising approach for achieving higher enhancements in both spectrum efficiency (SE) and energy efficiency (EE). Nevertheless, due to the double fading effect in reflecting links, the detection performance gains obtained through current passive RIS are negligible, especially in typical communication scenarios. As a result, the performance enhancement in passive RIS-assisted CSD is ineffective. To address the issue, this paper incorporates the advanced active RIS into CSD to assist the spectrum requestor (RU) in achieving notable and effective improvement in detection performance gains. Through amplifying the amplitude of reflected signals, active RIS can further enhance the received signal power at the RU while consuming more energy than the passive RIS. Therefore, for enabling both effective and efficient CSD with the assistance of active RIS, this paper investigates the detection efficiency (DE) maximization problem by jointly optimizing the RU’s sampling number and reward budget that is paid to compensate active RIS controllers for amplifying reflected signals. In a word, the key point of this paper is to evaluate whether the active RIS-assisted CSD can outperform the passive RIS-assisted CSD in both effective and efficient detection performance gains by joint parameter optimization. Quadratic transform (QT) and particle swarm optimization (PSO) methods are leveraged to solve the DE maximization problem. Finally, extensive simulation results demonstrate that, when taking the no-RIS mechanism as the benchmark, the proposed active RIS-assisted CSD can outperform the conventional passive-RIS mechanism in terms of significantly improved detection performance gains and a higher DE in most simulation scenarios.