The　last　few　years　have　seen　the　appearance　of　a　new　sparse　reference-based　free　energy　principle,　which　demonstrates　that　an　input　visual　signal　was　always　strived　to　be　comprehend　by　the　human　visual　system　(HVS)　through　removing　the　undetermined　portions.　By　this　inspiration,　we　in　this　paper　put　forward　an　image　quality　assessment　(IQA)　model　based　on　sparse　reference　information.　In　essential,　as　compared　with　the　classical　good-performance　FEDM　model　that　was　developed　with　a　linear　local　autoregressive　(AR)　model,　the　proposed　sparse　reference　(SR)　IQA　model　is　a　simplified　version　of　FEDM,　achieving　comparable　performance　but　hundreds　of　times　faster　implementation.　More　specifically,　this　paper　introduces　the　extremely　sparse　random　sampling　method　into　the　FEDM　model.　Experimental　results　on　the　most　well-known　LIVE　IQA　database　illustrate　the　effectiveness　along　with　efficiency　of　our　SR-IQA　model　as　compared　with　the　typical　full-reference　IQA　models　and　the　cutting-edge　SR　IQA　models.　©　2022,　Springer　Nature　Singapore　Pte　Ltd.