Use explicit UTF-8 encoding for icdar13_recognition annotation converter (#4062)

Author: pwolnows

tools/accuracy_checker/accuracy_checker/annotation_converters/icdar.py

@@ -178,7 +178,7 @@ def convert(self, check_content=False, progress_callback=None, progress_interval
             content_errors = []
             self.images_dir = self.images_dir or self.annotation_file.parent
 
-        original_annotations = read_txt(self.annotation_file)
+        original_annotations = read_txt(self.annotation_file, encoding='utf-8')
         num_iterations = len(original_annotations)
 
         for line_id, line in enumerate(original_annotations):