The luminous z = 0.286 quasar HE 0450–2958 is interacting with a companion galaxy at 6.5 kpc distance and the whole system radiates in the infrared at the level of an ultraluminous infrared galaxy ( ULIRG ) . A so far undetected host galaxy triggered the hypothesis of a mostly “ naked ” black hole ( BH ) ejected from the companion by three-body interaction . We present new HST/NICMOS 1.6 \mu m imaging data at 0 \farcs 1 resolution and VLT/VISIR 11.3 \mu m images at 0 \farcs 35 resolution that are for the first time resolving the system in the near- and mid-infrared . We combine these data with existing optical HST and CO maps . ( i ) At 1.6 \mu m we find an extension N-E of the quasar nucleus that is likely a part of the host galaxy , though not its main body . If true , a combination with upper limits on a main body co-centered with the quasar brackets the host galaxy luminosity to within a factor of \sim 4 and places HE 0450–2958 directly onto the M _ { \mathrm { BH } } - M _ { \mathrm { bulge } } -relation for nearby galaxies . ( ii ) A dust-free line of sight to the quasar suggests a low dust obscuration of the host galaxy , but the formal upper limit for star formation lies at 60 M _ { \odot } /yr . HE 0450–2958 is consistent with lying at the high-luminosity end of Narrow-Line Seyfert 1 Galaxies , and more exotic explanations like a “ naked quasar ” are unlikely . ( iii ) All 11.3 \mu m radiation in the system is emitted by the quasar nucleus . It has warm ULIRG-strength IR emission powered by black hole accretion and is radiating at super-Eddington rate , L / L _ { \mathrm { Edd } } = 6.2 ^ { +3.8 } _ { -1.8 } , or 12 M _ { \odot } /year . ( iv ) The companion galaxy is covered in optically thick dust and is not a collisional ring galaxy . It emits in the far infrared at ULIRG strength , powered by Arp220-like star formation ( strong starburst-like ) . An M82-like SED is ruled out . ( v ) With its black hole accretion rate HE 0450–2958 produces not enough new stars to maintain its position on the M _ { \mathrm { BH } } - M _ { \mathrm { bulge } } -relation , and star formation and black hole accretion are spatially disjoint . This relation can either only be maintained averaging over a longer timescale ( \lesssim 500 Myr ) and/or the bulge has to grow by redistribution of preexisting stars . ( vi ) Systems similar to HE 0450–2958 with spatially disjoint ULIRG-strength star formation and quasar activity might be common at high redshifts but at z < 0.43 we only find < 4 % ( 3/77 ) candidates for a similar configuration .