FRENKIE-CHIANG commited on
Commit
7cd47e9
·
verified ·
1 Parent(s): affe8ae

Upload app.py with huggingface_hub

Browse files
Files changed (1) hide show
  1. app.py +29 -23
app.py CHANGED
@@ -341,51 +341,57 @@ with gr.Blocks(
341
  outputs=[omni_table, image_table, video_table, audio_table],
342
  )
343
 
344
- gr.Markdown(
 
345
  """
346
  <div class="overall-definition">
347
-
348
  <h3>📊 Overall Score Definition</h3>
349
 
350
  <p>
351
- To facilitate clearer and more consistent comparison across models, we introduce an
352
- <b>Overall</b> score for each leaderboard track. The aggregation strategy is tailored
353
- to the evaluation protocol of each task category:
354
  </p>
355
 
356
  <p><b>1. OmniLLM / MLLM</b><br>
357
- The <b>Overall</b> score is computed as the arithmetic mean of all reported task-specific scores.
358
  </p>
359
 
360
  <p><b>2. Image Generation</b><br>
361
- The evaluation involves metrics defined on different numerical scales.
362
- <b>WIScore</b> is used for image generation, while <b>VIEScore</b> (averaged over three dimensions)
363
- is used for image editing.
364
  </p>
365
 
366
- <p>
367
- The <b>Overall</b> score is defined as:
368
  </p>
 
 
 
369
 
370
- <p style="text-align:center; font-size:16px;">
371
- \\[
372
- \\text{Overall} = \\frac{(\\text{WIScore} \\times 10) + \\left(\\frac{\\sum \\text{VIEScore}}{3}\\right)}{2}
373
- \\]
374
- </p>
 
 
375
 
 
 
 
376
  <p>
377
- This normalization-based formulation ensures a balanced contribution from both image generation
378
- and image editing performance.
379
  </p>
380
 
381
  <p><b>3. Video Generation</b><br>
382
- The <b>Overall</b> score is calculated as the arithmetic mean of all evaluated dimensions,
383
- including imaging quality, aesthetics, motion, and temporal consistency.
384
  </p>
385
-
386
  </div>
387
- """,
388
- unsafe_allow_html=True,
389
  )
390
 
391
  demo.launch()
 
341
  outputs=[omni_table, image_table, video_table, audio_table],
342
  )
343
 
344
+ # ---------- Overall definition (bottom) ----------
345
+ gr.HTML(
346
  """
347
  <div class="overall-definition">
 
348
  <h3>📊 Overall Score Definition</h3>
349
 
350
  <p>
351
+ To facilitate clearer and more consistent comparison across models, we introduce an
352
+ <b>Overall</b> score for each leaderboard track. The aggregation strategy is tailored
353
+ to the evaluation protocol of each task category:
354
  </p>
355
 
356
  <p><b>1. OmniLLM / MLLM</b><br>
357
+ The <b>Overall</b> score is computed as the arithmetic mean of all reported task-specific scores.
358
  </p>
359
 
360
  <p><b>2. Image Generation</b><br>
361
+ The evaluation involves metrics defined on different numerical scales.
362
+ <b>WIScore</b> is used for image generation, while <b>VIEScore</b> (averaged over three dimensions)
363
+ is used for image editing.
364
  </p>
365
 
366
+ <p style="margin-bottom: 6px;">
367
+ The <b>Overall</b> score is defined as:
368
  </p>
369
+ </div>
370
+ """
371
+ )
372
 
373
+ gr.Markdown(
374
+ r"""
375
+ \[
376
+ \text{Overall}=\frac{(\text{WIScore}\times 10)+\left(\frac{\sum \text{VIEScore}}{3}\right)}{2}
377
+ \]
378
+ """
379
+ )
380
 
381
+ gr.HTML(
382
+ """
383
+ <div class="overall-definition" style="margin-top: -24px;">
384
  <p>
385
+ This normalization-based formulation ensures a balanced contribution from both image generation
386
+ and image editing performance.
387
  </p>
388
 
389
  <p><b>3. Video Generation</b><br>
390
+ The <b>Overall</b> score is calculated as the arithmetic mean of all evaluated dimensions,
391
+ including imaging quality, aesthetics, motion, and temporal consistency.
392
  </p>
 
393
  </div>
394
+ """
 
395
  )
396
 
397
  demo.launch()