In this paper, spectrum transducing detection of infrared light with silicon detectors is systematic reviewed. Traditional infrared detection is based on semiconductor photonic detectors, such as AsGaIn and HgCdTe. These detectors have low detection sensitivities and relatively high noise at room temperature, and deep cooling is required to get better sensitivity. While the detection performances of silicon detectors are much better than those of the infrared detectors. Therefore, an effective method to detect infrared light is to transfer the wavelength of the infrared light to the detection window of silicon detector. Based on this principle, spectrum transducing detection of infrared light with silicon detectors is developed by using frequency up-conversion via sum frequency generation. This new detection scheme has the potential to offer single photon detection sensitivity at room temperature, which is very promising to be used in remote sensing at infrared regime.The main progress for spectrum transducing detection of infrared light can be divided into two groups. The first group is aimed at improving the key parameters in frequency conversion, which are quantum efficiency, noise, frequency bandwidth and spatial bandwidth. The conversion efficiency in frequency transducing can be enhanced by using cavity and waveguide (Fig.4), both configurations are demonstrated to achieve near unity internal conversion efficiencies; Noise in frequency conversion is mainly caused by spontaneous Raman scattering and parametric down conversion of strong pump beam, which can be measured at different pump configuration, and some effective methods can be used to sufficiently reduce the noise. These methods include: long wavelength pump laser, narrow band filters and reduction of the operation temperature of the nonlinear crystals. The frequency bandwidth is strongly dependent on the phase matching conditions, therefore effective methods such as chirped poling and multi-angle cut crystal can be used to enhance the frequency bandwidth in frequency conversion (Fig.5). The spatial bandwidth is dependent on crystal dimensions and phase matching, crystals with large optical aperture and large phase matching angles are preferred for spectrum transducing detection of image with large field of view, about 30 degree field of view is realized in mid-infrared up-conversion based on chirped PPLN crystal (Fig.6). The second groups of progresses aimed at applications of the spectrum transducing detection in different fields, these fields are: single photon detection at mid-infrared regime and quantum frequency interface (Fig.7) for applications in quantum information processing; classical optical imaging such as large field of view and high frame rate imaging in the mid-infrared regime, phase contrast imaging (Fig.8) and spectrum analysis for material sciences.For the mutual restrictions between different key parameters in spectrum transducing detection, one need to balance between different parameters for specific applications. Though the performance of spectrum transducing detection at the near infrared regime is high enough for some mentioned applications, the performances at mid-infrared is still not satisfied for typical applications, great efforts should be taken to improve the performance at this wavelength regime. For imaging detection based on spectrum transducing plane detectors, most studies are focused on coherent illumination, many key problems for illuminating with large bandwidth incoherent blackbody radiations are still not solved yet. In summaries, there are still opportunities for researches inspectrum transducing detection, these opportunities are: (1) to extending quantum optics and quantum spectroscopy to mid-infrared regime; (2) by combining spectrum transducing in interferometers to realize detection of infrared signal with undetected photons and optical phase amplification; (3) to transduce all other spectrums to the detection windows of silicon detectors and greatly reducing the detection complexity in large optical systems.