In optical microscopy, high-resolution (HR) volumetric imaging of thick biological specimens is highly desirable for many biomedical applications, such as development biology, tissue pathology, digital histology, and neuroscience. To obtain information on cellular events from the larger organism, e.g., a live embryo, intact tissue, or an organ, spatiotemporal patterns from the micro- to mesoscale must be in toto determined and analyzed.1–5 Thus, there is a growing need to develop HR, high-throughput imaging methods that can map entire large-volume specimens at high-spatiotemporal resolution.6,7 Recently, light-sheet microscopy (LSM) has emerged as a technique of choice that can image samples with low phototoxicity and at high speed.8–23 However, similar to conventional epifluorescence methods, LSM remains subject to the fundamental trade-off between high illumination/detection numerical apertures (NAs) and wide imaging fields of view (FOVs). In addition, an accurate digital sampling by the camera is also compromised by the need for large pixel size with high-fluorescence sensitivity. Therefore, the achievable resolution of current LSM systems is often pixel-limited under large FOVs, yielding inadequate optical throughput for digital imaging of mesoscale organisms at the cellular resolution. Tile imaging-based LSM systems have been developed to artificially increase the space-bandwidth product (SBP),24 hence, realizing HR imaging of large specimens.18,25–29 Despite the compromised speed induced by repetitive mechanical stitching, the high illumination/detection NA configuration in tile imaging induces increased phototoxicity for increasing sample size and limits fluorescence extraction from deep tissue. In addition, several techniques, such as Fourier ptychographic microscopy,30,31 synthetic aperture microscopy,32–35 contact-imaging microscopy,36,37 wavelength scanning microscopy,38 and lens-free digital holography,39–41 have recently provided a computational means of reconstructing a wide-FOV, HR image based on a number of low-resolution (LR) frames having certain correlations in the space, frequency, or spectrum domain.42–44 However, the majority of these methods target two-dimensional (2-D) bright-field microscopy and are not compatible with volumetric fluorescence imaging of thick samples.