[demo] Update demo & benchmarks

felixdittrich92 · Oct 22, 2024 · 143111b · 143111b
1 parent 7f63662
commit 143111b
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 10 deletions.
diff --git a/README.md b/README.md
@@ -321,15 +321,15 @@ Benchmarking performed on the FUNSD dataset and CORD dataset.
 
 docTR / OnnxTR models used for the benchmarks are `fast_base` (full precision) | `db_resnet50` (8-bit variant) for detection and `crnn_vgg16_bn` for recognition.
 
-The smallest combination in OnnxTR (docTR) of `db_mobilenet_v3_large` and `crnn_mobilenet_v3_small` takes as comparison `~0.17s / Page` on the FUNSD dataset and `~0.12s / Page` on the CORD dataset in **full precision**.
+The smallest combination in OnnxTR (docTR) of `db_mobilenet_v3_large` and `crnn_mobilenet_v3_small` takes as comparison `~0.17s / Page` on the FUNSD dataset and `~0.12s / Page` on the CORD dataset in **full precision** on CPU.
 
 - CPU benchmarks:
 
 |Library                          |FUNSD (199 pages)              |CORD  (900 pages)              |
 |---------------------------------|-------------------------------|-------------------------------|
 |docTR (CPU) - v0.8.1             | ~1.29s / Page                 | ~0.60s / Page                 |
-|**OnnxTR (CPU)** - v0.4.1        | ~0.57s / Page                 | **~0.25s / Page**             |
-|**OnnxTR (CPU) 8-bit** - v0.4.1  | **~0.38s / Page**             | **~0.14s / Page**             |
+|**OnnxTR (CPU)** - v0.5.1        | ~0.57s / Page                 | **~0.25s / Page**             |
+|**OnnxTR (CPU) 8-bit** - v0.5.1  | **~0.38s / Page**             | **~0.14s / Page**             |
 |EasyOCR (CPU) - v1.7.1           | ~1.96s / Page                 | ~1.75s / Page                 |
 |**PyTesseract (CPU)** - v0.3.10  | **~0.50s / Page**             | ~0.52s / Page                 |
 |Surya (line) (CPU) - v0.4.4      | ~48.76s / Page                | ~35.49s / Page                |
@@ -341,7 +341,8 @@ The smallest combination in OnnxTR (docTR) of `db_mobilenet_v3_large` and `crnn_
 |-------------------------------------|-------------------------------|-------------------------------|
 |docTR (GPU) - v0.8.1                 | ~0.07s / Page                 | ~0.05s / Page                 |
 |**docTR (GPU) float16** - v0.8.1     | **~0.06s / Page**             | **~0.03s / Page**             |
-|OnnxTR (GPU) - v0.4.1                | **~0.06s / Page**             | ~0.04s / Page                 |
+|OnnxTR (GPU) - v0.5.1                | **~0.06s / Page**             | ~0.04s / Page                 |
+|**OnnxTR (GPU) float16 - v0.5.1**    | **~0.05s / Page**             | **~0.03s / Page**             |
 |EasyOCR (GPU) - v1.7.1               | ~0.31s / Page                 | ~0.19s / Page                 |
 |Surya (GPU) float16 - v0.4.4         | ~3.70s / Page                 | ~2.81s / Page                 |
 |**PaddleOCR (GPU) - no cls - v2.7.3**| ~0.08s / Page                 | **~0.03s / Page**             |

diff --git a/demo/app.py b/demo/app.py
@@ -240,7 +240,7 @@ def analyze_page(
     )
     with gr.Row():
         with gr.Column(scale=1):
-            upload = gr.File(label="Upload File [JPG | PNG | PDF]", file_types=["pdf", "jpg", "png"])
+            upload = gr.File(label="Upload File [JPG | PNG | PDF]", file_types=[".pdf", ".jpg", ".png"])
             page_selection = gr.Slider(minimum=1, maximum=10, step=1, value=1, label="Page selection")
             det_model = gr.Dropdown(choices=DET_ARCHS, value=DET_ARCHS[0], label="Text detection model")
             reco_model = gr.Dropdown(
@@ -260,14 +260,14 @@ def analyze_page(
             analyze_button = gr.Button("Analyze page")
         with gr.Column(scale=3):
             with gr.Row():
-                input_image = gr.Image(label="Input page", width=600)
-                segmentation_heatmap = gr.Image(label="Segmentation heatmap", width=600)
-                output_image = gr.Image(label="Output page", width=600)
+                input_image = gr.Image(label="Input page", width=700, height=500)
+                segmentation_heatmap = gr.Image(label="Segmentation heatmap", width=700, height=500)
+                output_image = gr.Image(label="Output page", width=700, height=500)
             with gr.Row():
                 with gr.Column(scale=3):
-                    ocr_output = gr.JSON(label="OCR output", render=True, scale=1)
+                    ocr_output = gr.JSON(label="OCR output", render=True, scale=1, height=500)
                 with gr.Column(scale=3):
-                    synthesized_page = gr.Image(label="Synthesized page", width=600)
+                    synthesized_page = gr.Image(label="Synthesized page", width=700, height=500)
 
     analyze_button.click(
         analyze_page,