I will review Michiyama et al., (2020) (arXiv:2004.06123). A galaxy-galaxy merger is a fundamental process to understand galaxy evolution. It is said that a merger process triggers starburst activities. The classical idea is that starbursts are triggered in “nuclear” (inner central ~ 1 kpc) regions due to massive inflow. However, recent numerical simulations suggest the importance of “disk-wide” starbursts. It is necessary to observationally quantify the contribution of nuclear and disk-wide starburst activities. As a kickoff study, we selected one merging galaxy NGC 3256 which is one of the nearest infrared luminous galaxies in the universe. At first, we use MUSE on VLT and obtain maps of optical recombination lines (i.e., Hα and Hβ) in order to investigate the regions with modest dust extinction. We found many star-forming blobs (with the scale of ~100 pc) outside of nuclear regions. These regions are categorized as starburst in terms of surface density of star formation rate (SFR) and molecular gas mass. Second, we use mm/sub-mm recombination lines to investigate star-formation activity in very dusty regions. We observed H40α in 3mm by using ALMA. We found that there is a star-forming region where optical recombination lines are undetected due to strong dust extinction. Including both VLT and ALMA observation, we found that the contribution of SFR from nuclear and spatially extended starburst is ~34% and ~66%, respectively. Our observations clearly show the importance of spatially extended starbursts. These challenges the classical picture of “nuclear” starburst in a merger process.