An　efficient　mode　decision　scheme　for　down-sizing　video　transcoding　in　H.　264　using　support　vector　machines　(SVM)　was　proposed.　In　order　to　reduce　the　high　computational　complexity　of　using　conventional　mode　decision　in　the　H.　264　re-encoder,　the　proposed　scheme　used　SVM　to　exploit　the　correlation　between　coding　information　extracted　from　the　input　high-resolution　bit-stream　and　the　coding　modes　of　macro-blocks　in　down-sized　video　frames.　The　key　techniques　of　training　and　predicting　SVM　including　feature　vector　and　kernel　function　were　studied　and　then　the　SVM　model　was　trained.　After　the　hierarchical　SVM　classifier,　improbable　modes　were　eliminated　and　only　a　small　number　of　candidate　modes　were　carried　out　using　the　RDO　operations.　Hence,　remarkable　computing　time　could　be　saved,　up　to　67.31%,　while　maintaining　nearly　the　same　quality　of　the　transcoded　pictures.