Vocational　training　holds　significant　importance　in　the　Chinese　educational　system,　with　a　deliberate　emphasis　on　reform　and　innovation.　Enhancing　teaching　quality　is　a　priority,　and　conducting　a　comprehensive　assessment　of　it　is　essential.　Artificial　intelligence,　particularly　deep　learning　technology,　emerges　as　a　promising　solution　due　to　its　ability　to　effectively　handle　the　complex　and　diverse　aspects　involved　in　evaluating　teaching　quality　in　vocational　education.　In　this　manuscript　Chinese　vocational　skills　education　quality　assessment　using　attentive　dual　residual　generative　adversarial　network　optimised　with　gazelle　optimisation　algorithm　(CVSE-ADRGAN-GOA)　is　proposed.　Initially,　the　input　data　is　amassed　from　Chinese　vocational　skills　education　real　time　data.　The　acquired　data　is　preprocessed　using　the　colour　Wiener　filtering　method　for　normalising　the　input　data.　Then,　the　preprocessed　data　is　given　to　attentive　dual　residual　generative　adversarial　network　(ADRGAN)　for　the　quality　assessment　of　Chinese　vocational　skills　education.　The　gazelle　optimisation　algorithm　(GOA)　is　used　to　optimise　the　input　weight　parameters　of　the　ADRGAN.　The　proposed　CVSE-ADRGAN-GOA　technique　is　activated　in　MATLAB　and　its　efficacy　is　evaluated　utilising　some　performance　metrics,　like　accuracy,　precision,　sensitivity,　F1-score,　specificity,　error　rate,　receiver　operating　characteristic　(ROC),　computational　time.　The　proposed　CVSE-ADRGAN-GOA　method　provides　22.43%,　21.76%,　25.65%　higher　accuracy,　25.67%,　22.66%,　27.92%　higher　precision　while　compared　to　the　existing　models,　like　Internet　environment　and　machine　learning　espoused　innovation　way　of　secretarial　education　in　higher　vocational　schools　(CVSE-LRPM),　deep　learning　based　vocational　education　teaching　reform　quality　assessment　(CVSE-BPNN),　and　machine　learning　and　improved　support　vector　machine　(SVM)-based　teaching　reform　of　undergraduate　courses　in　colleges　and　universities　(CVSE-SVM)　methods,　respectively.　©　2024　Acta　Press.　All　rights　reserved.