Upload 13 files

sign2img.py

@@ -2,20 +2,46 @@ from io import BytesIO
 from signwriting.visualizer.visualize import signwriting_to_image
 from PIL import Image
 
-
 def to_bytes_io(image: Image) -> bytes:
     imgByte = BytesIO()
     image.save(imgByte, "png")
     imgByte.seek(0)
     return imgByte.getvalue()
 
-
-def convert(fsw: str, line_color=None, fill_color=None) -> bytes:
+def convert(fsw: str, line_color=None, fill_color=None, arrangement="row") -> bytes:
     if not line_color or len(line_color) != 4:
         line_color = (0, 0, 0, 255)
     
     if not fill_color or len(fill_color) != 4:
         fill_color = (0, 0, 0, 0)
-	
-    img: Image = signwriting_to_image(fsw, line_color=line_color, fill_color=fill_color)
-    return to_bytes_io(img)
+    
+    fsw_list = fsw.split()
+
+    # Convert each FSW string into an image and store them in a list
+    images = []
+    for fsw_string in fsw_list:
+        img = signwriting_to_image(fsw_string, line_color=line_color, fill_color=fill_color)
+        images.append(img)
+    
+    # Determine the maximum height and total width of all images
+    max_height = max(img.height for img in images)
+    total_width = sum(img.width for img in images)
+
+    # Create a new image to hold all images either row-wise or column-wise
+    if arrangement == "row":
+        final_image = Image.new("RGBA", (total_width, max_height), (255, 255, 255, 0))
+    elif arrangement == "column":
+        final_image = Image.new("RGBA", (max_height, total_width), (255, 255, 255, 0))
+    
+    # Paste each image onto the final image either row-wise or column-wise
+    x_offset = 0
+    y_offset = 0
+    for img in images:
+        if arrangement == "row":
+            final_image.paste(img, (x_offset, 0))
+            x_offset += img.width
+        elif arrangement == "column":
+            final_image.paste(img, (0, y_offset))
+            y_offset += img.height
+    
+    return to_bytes_io(final_image)

text2gloss.py

@@ -61,16 +61,34 @@ TRANSLATOR = Translator(
 
 def translate(text: str) -> str:
     spoken = text.lower().strip().split()
+
+    if all(c.isdigit() for c in spoken):
+        return text
+
     spoken.append(".")
 
     src_seq = [SRC.vocab.stoi.get(word, unk_idx) for word in spoken]
     pred_seq = TRANSLATOR.translate_sentence(
         torch.LongTensor([src_seq]).to(device))
+    pred_seq = set(pred_seq)
     pred_line = ' '.join(TRG.vocab.itos[idx] for idx in pred_seq)
     pred_line = pred_line.replace(Constants.BOS_WORD, '').replace(
         Constants.EOS_WORD, '').replace(Constants.PAD_WORD, '').replace(Constants.UNK_WORD, '')
 
     final = str(pred_line.strip())
+
+    if not contains_alpha_or_digits(final):
+        return text.lower().strip()
+
     if final.endswith(" ."):
         final = final[:-2]
+
+    if "i" in spoken and "me" not in final.split():
+        final = "me " + final
+
     return final
+
+def contains_alpha_or_digits(s: str) -> bool:
+    contains_alpha = any(c.isalpha() for c in s)
+    contains_digits = any(c.isdigit() for c in s)
+    return any([contains_alpha, contains_digits])