Supervisor's review for the bachelor's thesis of the 4th year student of the Department of Informatics of the Mathematics and Mechanics Department Rumyantsev V.F.  Use of machine learning methods for semantic classification of road conditions Bachelor's work of Rumyantsev is devoted to the problem of dividing the image from the camera on a vehicle into individual objects with the definition of their type. The task is actual, because the selection and classification of objects are actively used in various tasks of navigation, preventing collisions and controlling unmanned vehicles.  The aim of the work was to review the existing methods of semantic classification of the road situation and compare them to open databases.  The following results are achieved in the work: • Implemented and trained 3 neural network models SegNet, UNet, ENet; • Open databases of Cityscapes and CamVid are considered; • The accuracy and speed of the algorithms are compared on the considered databases; the best accuracy and at the same time the speed was demonstrated by the Enet algorithm, which has the largest depth of a network with an irregular structure and the least number of parameters.  The results are of great practical importance for the choice of the method of semantic classification of road conditions in the development of unmanned vehicles.  The work has also limitations: • Classification metrics for individual classes are not represented; • the text of the thesis note has some stylistic shortcomings.  Despite the shortcomings, given the valuable practical results of the work, I believe that the bachelor's thesis of Rumyantsev V.F. deserves an evaluation - "EXCELLENT".  PhD. senior lecturer of Department of Informatics Faculty of Mathematics and Mechanics of St. Petersburg State University Salischev S.I. 05/15/2018