Studies on visible-light-driven photocatalysis of air pollutants at indoor air quality (IAQ) levels have been limited. Current study investigated visible-light derived photocatalysis with N-doped and S-doped titanium dioxide ($TiO_2$) for the control of benzene at indoor levels. Two preparation processes were employed for each of the two types of photocatalyst: urea-Degussa P-25 $TiO_2$ and titania-colloid methods for the N-doped $TiO_2$; and titanium isopropoxid- and tetraisopropoxide-thiourea methods for the S-doped $TiO_2$. Furthermore, two coating methods (EDTA- and acetylacetone-dissolving methods) were tested for both the N-doped and S-doped $TiO_2$. The two coating methods exhibited different photocatalytic degradation efficiency for the N-doped photocatalysts, whereas they did not exhibit any difference for the S-doped photocatalysts. In addition, the two doping processes showed different photocatalytic degradation efficiency for both the S-doped and N-doped photocatalysts. For both the N-doped and S-doped $TiO_2$, the photocatalytic oxidation (PCO) efficiency increased as the hydraulic diameter (HD) decreased. The degradation efficiency determined via a PCO system with visible-light induced $TiO_2$ was lower than that with UV-light induced unmodified $TiO_2$, which was obtained from previous studies. Nevertheless, it is noteworthy that for the photocatalytic annular reactor with the HD of 0.5 cm, PCO efficiency increased up to 52% for the N-doped $TiO_2$ and 60% for the S-doped $TiO_2$. Consequently, when combined with the advantage of visible light use over UV light use, it is suggested that with appropriate HD conditions, the visible-light-assisted photocatalytic systems can also become an important tool for improving IAQ.