Spaces:
Sleeping
Source column + Delete button + absolute-index admin flow
Browse filesPublic leaderboard:
* New Src column shows whether each row was pre-populated by the
maintainer ('Admin', navy badge) or submitted via the public form
('User', purple badge). The submit handler stamps every uploaded
entry as submitter='user'; pre-populated rows arrive with
submitter='admin' from build_results.py in the Benchmark code repo.
Admin tab:
* admin_view now lists every entry (pending / approved / rejected)
with its absolute index and Source column; the per-status filtering
is gone so the # number on screen is stable.
* admin_act / admin_approve / admin_reject use absolute indices
directly. The new admin_delete pops the row from submissions.jsonl
permanently.
* Adds a ποΈ Delete button alongside Approve and Reject. The on-screen
hint explains that all three buttons take the absolute index from
the # column.
submissions.jsonl regenerated from build_results.py with submitter
fields on the eleven baselines.
- app.py +62 -19
- submissions.jsonl +11 -11
|
@@ -198,6 +198,8 @@ LEADERBOARD_CSS = """
|
|
| 198 |
.vbench-tbl .badge.D { background: #dadcff; color: #322a87; }
|
| 199 |
.vbench-tbl .badge.Ref { background: #e6e0ff; color: #3b1f7a; }
|
| 200 |
.vbench-tbl .badge.Other { background: #e3e8ee; color: #0a2540; }
|
|
|
|
|
|
|
| 201 |
|
| 202 |
.vbench-tbl .links a {
|
| 203 |
text-decoration: none; margin-right: 6px; font-size: 13px;
|
|
@@ -409,6 +411,7 @@ def _render_leaderboard_html_inner() -> str:
|
|
| 409 |
'<th class="rank">#</th>'
|
| 410 |
'<th class="left">Method</th>'
|
| 411 |
'<th class="left">Authors / Org</th>'
|
|
|
|
| 412 |
'<th class="score">TextScoreβ</th>'
|
| 413 |
'<th class="text">SeqAccβ</th><th class="text">CharAccβ</th><th class="text">TTSβ</th>'
|
| 414 |
"<th>Flk_fβ</th><th>Flk_cβ</th><th>Wp_fβ</th><th>Wp_cβ</th>"
|
|
@@ -429,10 +432,16 @@ def _render_leaderboard_html_inner() -> str:
|
|
| 429 |
)
|
| 430 |
ts = _record_text_score(r)
|
| 431 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 432 |
lines.append("<tr>")
|
| 433 |
lines.append(f'<td class="rank">{i}</td>')
|
| 434 |
lines.append(f'<td class="left method">{method_with_note}</td>')
|
| 435 |
lines.append(f'<td class="left org" title="{org}">{org}</td>')
|
|
|
|
| 436 |
lines.append(f'<td class="score">{_fmt(ts, ".4f")}</td>')
|
| 437 |
|
| 438 |
for k, fmt, dim, cls in [
|
|
@@ -503,6 +512,7 @@ def parse_eval_json(file_obj) -> Tuple[Optional[Dict], Optional[str]]:
|
|
| 503 |
metrics["n_clips"] = (
|
| 504 |
len(data["per_clip"]) if isinstance(data.get("per_clip"), dict) else None
|
| 505 |
)
|
|
|
|
| 506 |
return metrics, None
|
| 507 |
|
| 508 |
|
|
@@ -539,46 +549,52 @@ def submit_handler(method, organization, family, paper_url, code_url, contact_em
|
|
| 539 |
|
| 540 |
|
| 541 |
def admin_view(passphrase: str):
|
|
|
|
|
|
|
|
|
|
| 542 |
if not ADMIN_PASSPHRASE:
|
| 543 |
return pd.DataFrame(), "β Admin passphrase not configured on this Space."
|
| 544 |
if passphrase != ADMIN_PASSPHRASE:
|
| 545 |
return pd.DataFrame(), "β Wrong passphrase."
|
| 546 |
items = fetch_submissions()
|
| 547 |
-
|
| 548 |
-
|
| 549 |
-
return pd.DataFrame(), "β
No pending submissions."
|
| 550 |
rows = []
|
| 551 |
-
for i, r in enumerate(
|
| 552 |
rows.append({
|
| 553 |
"#": i,
|
| 554 |
"Method": r.get("method", "β"),
|
|
|
|
|
|
|
|
|
|
| 555 |
"Family": r.get("family", "β"),
|
| 556 |
"Organization": r.get("organization", "β"),
|
| 557 |
-
"TextScore": _record_text_score(r),
|
| 558 |
-
"SeqAcc": r.get("SeqAcc"),
|
| 559 |
-
"CharAcc": r.get("CharAcc"),
|
| 560 |
-
"TTS": r.get("TTS"),
|
| 561 |
"Email": r.get("contact_email", ""),
|
| 562 |
"Submitted": r.get("submitted_at", ""),
|
| 563 |
})
|
| 564 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 565 |
|
| 566 |
|
| 567 |
def admin_act(passphrase: str, idx: int, action: str):
|
|
|
|
| 568 |
if not ADMIN_PASSPHRASE or passphrase != ADMIN_PASSPHRASE:
|
| 569 |
return "β Wrong passphrase.", render_leaderboard_html()
|
| 570 |
items = fetch_submissions()
|
| 571 |
-
|
| 572 |
-
|
| 573 |
-
|
| 574 |
-
target = pending_pos[int(idx)]
|
| 575 |
items[target]["status"] = action
|
| 576 |
items[target][f"{action}_at"] = time.strftime("%Y-%m-%d %H:%M:%S UTC", time.gmtime())
|
| 577 |
try:
|
| 578 |
write_submissions(items)
|
| 579 |
except Exception as e:
|
| 580 |
return f"β Could not save: {e}", render_leaderboard_html()
|
| 581 |
-
|
|
|
|
| 582 |
|
| 583 |
|
| 584 |
def admin_approve(passphrase: str, idx: int):
|
|
@@ -589,6 +605,22 @@ def admin_reject(passphrase: str, idx: int):
|
|
| 589 |
return admin_act(passphrase, idx, "rejected")
|
| 590 |
|
| 591 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 592 |
# ---------- UI ----------
|
| 593 |
|
| 594 |
HEADER_MD = (
|
|
@@ -666,16 +698,22 @@ with gr.Blocks(title="ViTeX-Bench Leaderboard",
|
|
| 666 |
)
|
| 667 |
admin_pass = gr.Textbox(label="Passphrase", type="password")
|
| 668 |
with gr.Row():
|
| 669 |
-
view_btn = gr.Button("Show
|
| 670 |
admin_msg = gr.Markdown()
|
| 671 |
-
|
| 672 |
-
view_btn.click(admin_view, inputs=admin_pass, outputs=[
|
| 673 |
|
| 674 |
-
gr.Markdown("---\n### Approve /
|
| 675 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 676 |
with gr.Row():
|
| 677 |
approve_btn = gr.Button("β
Approve", variant="primary")
|
| 678 |
reject_btn = gr.Button("β Reject")
|
|
|
|
| 679 |
action_msg = gr.Markdown()
|
| 680 |
approve_btn.click(
|
| 681 |
admin_approve,
|
|
@@ -687,6 +725,11 @@ with gr.Blocks(title="ViTeX-Bench Leaderboard",
|
|
| 687 |
inputs=[admin_pass, idx_in],
|
| 688 |
outputs=[action_msg, leaderboard_html],
|
| 689 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 690 |
|
| 691 |
with gr.Tab("βΉοΈ About"):
|
| 692 |
gr.Markdown(
|
|
|
|
| 198 |
.vbench-tbl .badge.D { background: #dadcff; color: #322a87; }
|
| 199 |
.vbench-tbl .badge.Ref { background: #e6e0ff; color: #3b1f7a; }
|
| 200 |
.vbench-tbl .badge.Other { background: #e3e8ee; color: #0a2540; }
|
| 201 |
+
.vbench-tbl .badge.Admin { background: #0a2540; color: #f6f9fc; }
|
| 202 |
+
.vbench-tbl .badge.User { background: #635bff; color: #fff; }
|
| 203 |
|
| 204 |
.vbench-tbl .links a {
|
| 205 |
text-decoration: none; margin-right: 6px; font-size: 13px;
|
|
|
|
| 411 |
'<th class="rank">#</th>'
|
| 412 |
'<th class="left">Method</th>'
|
| 413 |
'<th class="left">Authors / Org</th>'
|
| 414 |
+
'<th class="left">Src</th>'
|
| 415 |
'<th class="score">TextScoreβ</th>'
|
| 416 |
'<th class="text">SeqAccβ</th><th class="text">CharAccβ</th><th class="text">TTSβ</th>'
|
| 417 |
"<th>Flk_fβ</th><th>Flk_cβ</th><th>Wp_fβ</th><th>Wp_cβ</th>"
|
|
|
|
| 432 |
)
|
| 433 |
ts = _record_text_score(r)
|
| 434 |
|
| 435 |
+
submitter = (r.get("submitter") or "user").strip().lower()
|
| 436 |
+
src_label = "Admin" if submitter == "admin" else "User"
|
| 437 |
+
src_class = "Admin" if submitter == "admin" else "User"
|
| 438 |
+
src_badge = f'<span class="badge {src_class}">{src_label}</span>'
|
| 439 |
+
|
| 440 |
lines.append("<tr>")
|
| 441 |
lines.append(f'<td class="rank">{i}</td>')
|
| 442 |
lines.append(f'<td class="left method">{method_with_note}</td>')
|
| 443 |
lines.append(f'<td class="left org" title="{org}">{org}</td>')
|
| 444 |
+
lines.append(f'<td class="left">{src_badge}</td>')
|
| 445 |
lines.append(f'<td class="score">{_fmt(ts, ".4f")}</td>')
|
| 446 |
|
| 447 |
for k, fmt, dim, cls in [
|
|
|
|
| 512 |
metrics["n_clips"] = (
|
| 513 |
len(data["per_clip"]) if isinstance(data.get("per_clip"), dict) else None
|
| 514 |
)
|
| 515 |
+
metrics["submitter"] = "user"
|
| 516 |
return metrics, None
|
| 517 |
|
| 518 |
|
|
|
|
| 549 |
|
| 550 |
|
| 551 |
def admin_view(passphrase: str):
|
| 552 |
+
"""Show every entry (pending / approved / rejected) with its absolute
|
| 553 |
+
index so the admin operates on stable row numbers regardless of
|
| 554 |
+
status."""
|
| 555 |
if not ADMIN_PASSPHRASE:
|
| 556 |
return pd.DataFrame(), "β Admin passphrase not configured on this Space."
|
| 557 |
if passphrase != ADMIN_PASSPHRASE:
|
| 558 |
return pd.DataFrame(), "β Wrong passphrase."
|
| 559 |
items = fetch_submissions()
|
| 560 |
+
if not items:
|
| 561 |
+
return pd.DataFrame(), "π No submissions yet."
|
|
|
|
| 562 |
rows = []
|
| 563 |
+
for i, r in enumerate(items):
|
| 564 |
rows.append({
|
| 565 |
"#": i,
|
| 566 |
"Method": r.get("method", "β"),
|
| 567 |
+
"Status": r.get("status", "β"),
|
| 568 |
+
"Source": (r.get("submitter") or "β"),
|
| 569 |
+
"TextScore": _record_text_score(r),
|
| 570 |
"Family": r.get("family", "β"),
|
| 571 |
"Organization": r.get("organization", "β"),
|
|
|
|
|
|
|
|
|
|
|
|
|
| 572 |
"Email": r.get("contact_email", ""),
|
| 573 |
"Submitted": r.get("submitted_at", ""),
|
| 574 |
})
|
| 575 |
+
n_pending = sum(1 for r in items if r.get("status") == "pending")
|
| 576 |
+
return pd.DataFrame(rows), (
|
| 577 |
+
f"π {len(items)} entry(ies) β {n_pending} pending. "
|
| 578 |
+
"Use the absolute index in the # column for every action below."
|
| 579 |
+
)
|
| 580 |
|
| 581 |
|
| 582 |
def admin_act(passphrase: str, idx: int, action: str):
|
| 583 |
+
"""Approve or reject the entry at absolute index `idx`."""
|
| 584 |
if not ADMIN_PASSPHRASE or passphrase != ADMIN_PASSPHRASE:
|
| 585 |
return "β Wrong passphrase.", render_leaderboard_html()
|
| 586 |
items = fetch_submissions()
|
| 587 |
+
if idx is None or int(idx) < 0 or int(idx) >= len(items):
|
| 588 |
+
return f"β Index {idx} out of range ({len(items)} entries).", render_leaderboard_html()
|
| 589 |
+
target = int(idx)
|
|
|
|
| 590 |
items[target]["status"] = action
|
| 591 |
items[target][f"{action}_at"] = time.strftime("%Y-%m-%d %H:%M:%S UTC", time.gmtime())
|
| 592 |
try:
|
| 593 |
write_submissions(items)
|
| 594 |
except Exception as e:
|
| 595 |
return f"β Could not save: {e}", render_leaderboard_html()
|
| 596 |
+
method = items[target].get("method", "β")
|
| 597 |
+
return f"β
Entry #{target} ({method}) β {action}.", render_leaderboard_html()
|
| 598 |
|
| 599 |
|
| 600 |
def admin_approve(passphrase: str, idx: int):
|
|
|
|
| 605 |
return admin_act(passphrase, idx, "rejected")
|
| 606 |
|
| 607 |
|
| 608 |
+
def admin_delete(passphrase: str, idx: int):
|
| 609 |
+
"""Permanently remove the entry at absolute index `idx`."""
|
| 610 |
+
if not ADMIN_PASSPHRASE or passphrase != ADMIN_PASSPHRASE:
|
| 611 |
+
return "β Wrong passphrase.", render_leaderboard_html()
|
| 612 |
+
items = fetch_submissions()
|
| 613 |
+
if idx is None or int(idx) < 0 or int(idx) >= len(items):
|
| 614 |
+
return f"β Index {idx} out of range ({len(items)} entries).", render_leaderboard_html()
|
| 615 |
+
removed = items.pop(int(idx))
|
| 616 |
+
try:
|
| 617 |
+
write_submissions(items)
|
| 618 |
+
except Exception as e:
|
| 619 |
+
return f"β Could not save: {e}", render_leaderboard_html()
|
| 620 |
+
return (f"ποΈ Deleted #{int(idx)} ({removed.get('method', 'β')}).",
|
| 621 |
+
render_leaderboard_html())
|
| 622 |
+
|
| 623 |
+
|
| 624 |
# ---------- UI ----------
|
| 625 |
|
| 626 |
HEADER_MD = (
|
|
|
|
| 698 |
)
|
| 699 |
admin_pass = gr.Textbox(label="Passphrase", type="password")
|
| 700 |
with gr.Row():
|
| 701 |
+
view_btn = gr.Button("π Show all entries")
|
| 702 |
admin_msg = gr.Markdown()
|
| 703 |
+
all_df = gr.Dataframe(interactive=False, wrap=True)
|
| 704 |
+
view_btn.click(admin_view, inputs=admin_pass, outputs=[all_df, admin_msg])
|
| 705 |
|
| 706 |
+
gr.Markdown("---\n### Approve / Reject / Delete")
|
| 707 |
+
gr.Markdown(
|
| 708 |
+
"Enter the **absolute** row index from the `#` column above. "
|
| 709 |
+
"Approve / Reject set the entry's status; Delete removes the entry "
|
| 710 |
+
"permanently from `submissions.jsonl`."
|
| 711 |
+
)
|
| 712 |
+
idx_in = gr.Number(label="Row index", value=0, precision=0)
|
| 713 |
with gr.Row():
|
| 714 |
approve_btn = gr.Button("β
Approve", variant="primary")
|
| 715 |
reject_btn = gr.Button("β Reject")
|
| 716 |
+
delete_btn = gr.Button("ποΈ Delete", variant="stop")
|
| 717 |
action_msg = gr.Markdown()
|
| 718 |
approve_btn.click(
|
| 719 |
admin_approve,
|
|
|
|
| 725 |
inputs=[admin_pass, idx_in],
|
| 726 |
outputs=[action_msg, leaderboard_html],
|
| 727 |
)
|
| 728 |
+
delete_btn.click(
|
| 729 |
+
admin_delete,
|
| 730 |
+
inputs=[admin_pass, idx_in],
|
| 731 |
+
outputs=[action_msg, leaderboard_html],
|
| 732 |
+
)
|
| 733 |
|
| 734 |
with gr.Tab("βΉοΈ About"):
|
| 735 |
gr.Markdown(
|
|
@@ -1,11 +1,11 @@
|
|
| 1 |
-
{"method": "TextCtrl", "family": "A β per-frame image editor", "organization": "Zeng et al., 2024", "paper_url": "https://arxiv.org/abs/2410.10133", "code_url": "https://github.com/weichaozeng/TextCtrl", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.5623872104295862, "SeqAcc": 0.47475474732914913, "CharAcc": 0.733502509720617, "TTS": 0.5107817672969821, "Flicker_full": 3.8040257787170293, "Flicker_crop": 4.287049075959627, "Warp_full": 1.5883564410079873, "Warp_crop": 2.087607682354121, "MUSIQ_full": 70.32216657276115, "MUSIQ_crop": 42.77286880553454, "PSNR_loc": 41.143448625451185, "SSIM_loc": 0.9944056776770688, "LPIPS_loc": 0.007969770003940647, "DreamSim_loc": 0.0042883308893049855, "n_clips": 157}
|
| 2 |
-
{"method": "ViTeX-14B (Composite)", "family": "Reference", "organization": "Anonymous (NeurIPS 2026 D&B submission)", "paper_url": "", "code_url": "https://huggingface.co/ViTeX-Bench/ViTeX-14B", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.5409733932921607, "SeqAcc": 0.34489601685376864, "CharAcc": 0.6892114595238374, "TTS": 0.6660196589960885, "Flicker_full": 3.730421558994263, "Flicker_crop": 3.826727295895001, "Warp_full": 1.5060892347506618, "Warp_crop": 1.5591366257464094, "MUSIQ_full": 70.27118598215141, "MUSIQ_crop": 44.944762801223376, "PSNR_loc": 42.950774276028774, "SSIM_loc": 0.9925085173386224, "LPIPS_loc": 0.005916571278483934, "DreamSim_loc": 0.0023257043362929306, "n_clips": 157}
|
| 3 |
-
{"method": "ViTeX-14B", "family": "Reference", "organization": "Anonymous (NeurIPS 2026 D&B submission)", "paper_url": "", "code_url": "https://huggingface.co/ViTeX-Bench/ViTeX-14B", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.5337670099905598, "SeqAcc": 0.34121246792876553, "CharAcc": 0.6879770723988642, "TTS": 0.6478229475144797, "Flicker_full": 3.2739301912212606, "Flicker_crop": 3.424670762474799, "Warp_full": 1.5515188207705402, "Warp_crop": 1.5304199630586097, "MUSIQ_full": 69.63500067777694, "MUSIQ_crop": 43.52961571422055, "PSNR_loc": 29.077432591849323, "SSIM_loc": 0.9512201399006257, "LPIPS_loc": 0.06030903690911814, "DreamSim_loc": 0.023522706465862867, "n_clips": 157}
|
| 4 |
-
{"method": "VideoPainter", "family": "C β mask-conditioned video inpainting", "organization": "Bian et al., 2025", "paper_url": "https://arxiv.org/abs/2503.05639", "code_url": "https://github.com/TencentARC/VideoPainter", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.51506756458757, "SeqAcc": 0.364495972867382, "CharAcc": 0.6187952902754302, "TTS": 0.6058329243574021, "Flicker_full": 2.383399970485585, "Flicker_crop": 2.619418716169186, "Warp_full": 2.9276182078188366, "Warp_crop": 3.3452600061138558, "MUSIQ_full": 67.16001260609637, "MUSIQ_crop": 40.58771384010968, "PSNR_loc": 28.555957743164843, "SSIM_loc": 0.9151628155450829, "LPIPS_loc": 0.10402342236567201, "DreamSim_loc": 0.023908750937496278, "n_clips": 157}
|
| 5 |
-
{"method": "FLUX-Text", "family": "A β per-frame image editor", "organization": "Chen et al., 2025", "paper_url": "https://arxiv.org/abs/2505.03329", "code_url": "https://github.com/AMAP-ML/FluxText", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.5022999045945803, "SeqAcc": 0.5283744349135131, "CharAcc": 0.7367738685630717, "TTS": 0.32554668434702094, "Flicker_full": 5.114334507312302, "Flicker_crop": 14.81406893996351, "Warp_full": 3.0267581734528144, "Warp_crop": 13.009849474748862, "MUSIQ_full": 70.25921666161523, "MUSIQ_crop": 43.85439727157533, "PSNR_loc": 31.488873457756767, "SSIM_loc": 0.974685608615182, "LPIPS_loc": 0.028573400793733536, "DreamSim_loc": 0.012038936603600812, "n_clips": 157}
|
| 6 |
-
{"method": "RS-STE", "family": "A β per-frame image editor", "organization": "Zhao et al., 2025", "paper_url": "https://arxiv.org/abs/2503.17774", "code_url": "https://github.com/honglei-zhao/RS-STE", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.4907994847015915, "SeqAcc": 0.3539735290248826, "CharAcc": 0.6258597181299173, "TTS": 0.5336598236730271, "Flicker_full": 3.728183996539611, "Flicker_crop": 3.66286053942277, "Warp_full": 1.6050723235894067, "Warp_crop": 1.8147908492065754, "MUSIQ_full": 69.57172569297175, "MUSIQ_crop": 34.26484699258097, "PSNR_loc": 37.00242438437832, "SSIM_loc": 0.9830883838061237, "LPIPS_loc": 0.02354780357549038, "DreamSim_loc": 0.007322213357421243, "n_clips": 157}
|
| 7 |
-
{"method": "AnyText2", "family": "A β per-frame image editor", "organization": "Tuo et al., 2024", "paper_url": "https://arxiv.org/abs/2411.15245", "code_url": "https://github.com/tyxsspa/AnyText2", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.4074467792171563, "SeqAcc": 0.27973291436891057, "CharAcc": 0.6332210144114335, "TTS": 0.38186844987155294, "Flicker_full": 3.3398760175596904, "Flicker_crop": 4.9545532378085495, "Warp_full": 2.0425352598531266, "Warp_crop": 3.9516436691464083, "MUSIQ_full": 66.67552189796594, "MUSIQ_crop": 41.65317273156116, "PSNR_loc": 25.55582150532182, "SSIM_loc": 0.9047352040975998, "LPIPS_loc": 0.09148503486230188, "DreamSim_loc": 0.0430956823557552, "n_clips": 157}
|
| 8 |
-
{"method": "TextCtrl + AnyV2V", "family": "B β first-frame + I2V propagation", "organization": "Composite of Zeng 2024 + Ku 2024", "paper_url": "", "code_url": "", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.16492592568061926, "SeqAcc": 0.056786779447094926, "CharAcc": 0.30779952787986514, "TTS": 0.2566561069340611, "Flicker_full": 4.9802852382134075, "Flicker_crop": 4.977045501548225, "Warp_full": 4.107769233013561, "Warp_crop": 3.9674498647429868, "MUSIQ_full": 69.41088303841349, "MUSIQ_crop": 33.85112622206451, "PSNR_loc": 21.084345629665755, "SSIM_loc": 0.7846697544754223, "LPIPS_loc": 0.2249758477633198, "DreamSim_loc": 0.07321823651876672, "n_clips": 157}
|
| 9 |
-
{"method": "Identity (sanity)", "family": "β", "organization": "β", "paper_url": "", "code_url": "", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.0, "SeqAcc": 0.0, "CharAcc": 0.3165408461004333, "TTS": 0.7598748923514798, "Flicker_full": 3.7225515203842283, "Flicker_crop": 3.6811893970338505, "Warp_full": 1.46410686887435, "Warp_crop": 1.2690219210512326, "MUSIQ_full": 70.32922646265375, "MUSIQ_crop": 45.12224410273406, "PSNR_loc": 100.0, "SSIM_loc": 1.0, "LPIPS_loc": 0.0, "DreamSim_loc": -4.1919402509231723e-08, "n_clips": 157}
|
| 10 |
-
{"method": "Wan2.1-VACE-14B", "family": "C β mask-conditioned video inpainting", "organization": "Wan-AI, 2025", "paper_url": "https://arxiv.org/abs/2503.07598", "code_url": "https://huggingface.co/Wan-AI/Wan2.1-VACE-14B", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.0, "SeqAcc": 0.0, "CharAcc": 0.29842756354252414, "TTS": 0.6894925190282424, "Flicker_full": 3.777380530928976, "Flicker_crop": 3.841096694996713, "Warp_full": 1.6876533062098615, "Warp_crop": 1.5609657061512832, "MUSIQ_full": 70.53707020378923, "MUSIQ_crop": 45.256712742544, "PSNR_loc": 35.21163969961195, "SSIM_loc": 0.9761949368626082, "LPIPS_loc": 0.021842662243107273, "DreamSim_loc": 0.007056991990078281, "n_clips": 157}
|
| 11 |
-
{"method": "Kling Video 3.0 Omni", "family": "D β instruction-guided V2V", "organization": "Kuaishou (closed)", "paper_url": "", "code_url": "", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.0, "SeqAcc": 0.0, "CharAcc": 0.20752862556378657, "TTS": 0.6408892575809407, "Flicker_full": 4.247679066116782, "Flicker_crop": 4.081252510324976, "Warp_full": 3.1189079573144176, "Warp_crop": 2.902087520422552, "MUSIQ_full": 72.23268125973436, "MUSIQ_crop": 47.745845725613485, "PSNR_loc": 21.181631575850673, "SSIM_loc": 0.843030764793486, "LPIPS_loc": 0.17594784468030775, "DreamSim_loc": 0.060776721627595835, "n_clips": 157}
|
|
|
|
| 1 |
+
{"method": "TextCtrl", "family": "A β per-frame image editor", "organization": "Zeng et al., 2024", "paper_url": "https://arxiv.org/abs/2410.10133", "code_url": "https://github.com/weichaozeng/TextCtrl", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.5623872104295862, "SeqAcc": 0.47475474732914913, "CharAcc": 0.733502509720617, "TTS": 0.5107817672969821, "Flicker_full": 3.8040257787170293, "Flicker_crop": 4.287049075959627, "Warp_full": 1.5883564410079873, "Warp_crop": 2.087607682354121, "MUSIQ_full": 70.32216657276115, "MUSIQ_crop": 42.77286880553454, "PSNR_loc": 41.143448625451185, "SSIM_loc": 0.9944056776770688, "LPIPS_loc": 0.007969770003940647, "DreamSim_loc": 0.0042883308893049855, "n_clips": 157}
|
| 2 |
+
{"method": "ViTeX-14B (Composite)", "family": "Reference", "organization": "Anonymous (NeurIPS 2026 D&B submission)", "paper_url": "", "code_url": "https://huggingface.co/ViTeX-Bench/ViTeX-14B", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.5409733932921607, "SeqAcc": 0.34489601685376864, "CharAcc": 0.6892114595238374, "TTS": 0.6660196589960885, "Flicker_full": 3.730421558994263, "Flicker_crop": 3.826727295895001, "Warp_full": 1.5060892347506618, "Warp_crop": 1.5591366257464094, "MUSIQ_full": 70.27118598215141, "MUSIQ_crop": 44.944762801223376, "PSNR_loc": 42.950774276028774, "SSIM_loc": 0.9925085173386224, "LPIPS_loc": 0.005916571278483934, "DreamSim_loc": 0.0023257043362929306, "n_clips": 157}
|
| 3 |
+
{"method": "ViTeX-14B", "family": "Reference", "organization": "Anonymous (NeurIPS 2026 D&B submission)", "paper_url": "", "code_url": "https://huggingface.co/ViTeX-Bench/ViTeX-14B", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.5337670099905598, "SeqAcc": 0.34121246792876553, "CharAcc": 0.6879770723988642, "TTS": 0.6478229475144797, "Flicker_full": 3.2739301912212606, "Flicker_crop": 3.424670762474799, "Warp_full": 1.5515188207705402, "Warp_crop": 1.5304199630586097, "MUSIQ_full": 69.63500067777694, "MUSIQ_crop": 43.52961571422055, "PSNR_loc": 29.077432591849323, "SSIM_loc": 0.9512201399006257, "LPIPS_loc": 0.06030903690911814, "DreamSim_loc": 0.023522706465862867, "n_clips": 157}
|
| 4 |
+
{"method": "VideoPainter", "family": "C β mask-conditioned video inpainting", "organization": "Bian et al., 2025", "paper_url": "https://arxiv.org/abs/2503.05639", "code_url": "https://github.com/TencentARC/VideoPainter", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.51506756458757, "SeqAcc": 0.364495972867382, "CharAcc": 0.6187952902754302, "TTS": 0.6058329243574021, "Flicker_full": 2.383399970485585, "Flicker_crop": 2.619418716169186, "Warp_full": 2.9276182078188366, "Warp_crop": 3.3452600061138558, "MUSIQ_full": 67.16001260609637, "MUSIQ_crop": 40.58771384010968, "PSNR_loc": 28.555957743164843, "SSIM_loc": 0.9151628155450829, "LPIPS_loc": 0.10402342236567201, "DreamSim_loc": 0.023908750937496278, "n_clips": 157}
|
| 5 |
+
{"method": "FLUX-Text", "family": "A β per-frame image editor", "organization": "Chen et al., 2025", "paper_url": "https://arxiv.org/abs/2505.03329", "code_url": "https://github.com/AMAP-ML/FluxText", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.5022999045945803, "SeqAcc": 0.5283744349135131, "CharAcc": 0.7367738685630717, "TTS": 0.32554668434702094, "Flicker_full": 5.114334507312302, "Flicker_crop": 14.81406893996351, "Warp_full": 3.0267581734528144, "Warp_crop": 13.009849474748862, "MUSIQ_full": 70.25921666161523, "MUSIQ_crop": 43.85439727157533, "PSNR_loc": 31.488873457756767, "SSIM_loc": 0.974685608615182, "LPIPS_loc": 0.028573400793733536, "DreamSim_loc": 0.012038936603600812, "n_clips": 157}
|
| 6 |
+
{"method": "RS-STE", "family": "A β per-frame image editor", "organization": "Zhao et al., 2025", "paper_url": "https://arxiv.org/abs/2503.17774", "code_url": "https://github.com/honglei-zhao/RS-STE", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.4907994847015915, "SeqAcc": 0.3539735290248826, "CharAcc": 0.6258597181299173, "TTS": 0.5336598236730271, "Flicker_full": 3.728183996539611, "Flicker_crop": 3.66286053942277, "Warp_full": 1.6050723235894067, "Warp_crop": 1.8147908492065754, "MUSIQ_full": 69.57172569297175, "MUSIQ_crop": 34.26484699258097, "PSNR_loc": 37.00242438437832, "SSIM_loc": 0.9830883838061237, "LPIPS_loc": 0.02354780357549038, "DreamSim_loc": 0.007322213357421243, "n_clips": 157}
|
| 7 |
+
{"method": "AnyText2", "family": "A β per-frame image editor", "organization": "Tuo et al., 2024", "paper_url": "https://arxiv.org/abs/2411.15245", "code_url": "https://github.com/tyxsspa/AnyText2", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.4074467792171563, "SeqAcc": 0.27973291436891057, "CharAcc": 0.6332210144114335, "TTS": 0.38186844987155294, "Flicker_full": 3.3398760175596904, "Flicker_crop": 4.9545532378085495, "Warp_full": 2.0425352598531266, "Warp_crop": 3.9516436691464083, "MUSIQ_full": 66.67552189796594, "MUSIQ_crop": 41.65317273156116, "PSNR_loc": 25.55582150532182, "SSIM_loc": 0.9047352040975998, "LPIPS_loc": 0.09148503486230188, "DreamSim_loc": 0.0430956823557552, "n_clips": 157}
|
| 8 |
+
{"method": "TextCtrl + AnyV2V", "family": "B β first-frame + I2V propagation", "organization": "Composite of Zeng 2024 + Ku 2024", "paper_url": "", "code_url": "", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.16492592568061926, "SeqAcc": 0.056786779447094926, "CharAcc": 0.30779952787986514, "TTS": 0.2566561069340611, "Flicker_full": 4.9802852382134075, "Flicker_crop": 4.977045501548225, "Warp_full": 4.107769233013561, "Warp_crop": 3.9674498647429868, "MUSIQ_full": 69.41088303841349, "MUSIQ_crop": 33.85112622206451, "PSNR_loc": 21.084345629665755, "SSIM_loc": 0.7846697544754223, "LPIPS_loc": 0.2249758477633198, "DreamSim_loc": 0.07321823651876672, "n_clips": 157}
|
| 9 |
+
{"method": "Identity (sanity)", "family": "β", "organization": "β", "paper_url": "", "code_url": "", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.0, "SeqAcc": 0.0, "CharAcc": 0.3165408461004333, "TTS": 0.7598748923514798, "Flicker_full": 3.7225515203842283, "Flicker_crop": 3.6811893970338505, "Warp_full": 1.46410686887435, "Warp_crop": 1.2690219210512326, "MUSIQ_full": 70.32922646265375, "MUSIQ_crop": 45.12224410273406, "PSNR_loc": 100.0, "SSIM_loc": 1.0, "LPIPS_loc": 0.0, "DreamSim_loc": -4.1919402509231723e-08, "n_clips": 157}
|
| 10 |
+
{"method": "Wan2.1-VACE-14B", "family": "C β mask-conditioned video inpainting", "organization": "Wan-AI, 2025", "paper_url": "https://arxiv.org/abs/2503.07598", "code_url": "https://huggingface.co/Wan-AI/Wan2.1-VACE-14B", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.0, "SeqAcc": 0.0, "CharAcc": 0.29842756354252414, "TTS": 0.6894925190282424, "Flicker_full": 3.777380530928976, "Flicker_crop": 3.841096694996713, "Warp_full": 1.6876533062098615, "Warp_crop": 1.5609657061512832, "MUSIQ_full": 70.53707020378923, "MUSIQ_crop": 45.256712742544, "PSNR_loc": 35.21163969961195, "SSIM_loc": 0.9761949368626082, "LPIPS_loc": 0.021842662243107273, "DreamSim_loc": 0.007056991990078281, "n_clips": 157}
|
| 11 |
+
{"method": "Kling Video 3.0 Omni", "family": "D β instruction-guided V2V", "organization": "Kuaishou (closed)", "paper_url": "", "code_url": "", "submitter": "admin", "submitted_at": "2026-05-04 00:00:00 UTC", "approved_at": "2026-05-04 00:00:00 UTC", "status": "approved", "TextScore": 0.0, "SeqAcc": 0.0, "CharAcc": 0.20752862556378657, "TTS": 0.6408892575809407, "Flicker_full": 4.247679066116782, "Flicker_crop": 4.081252510324976, "Warp_full": 3.1189079573144176, "Warp_crop": 2.902087520422552, "MUSIQ_full": 72.23268125973436, "MUSIQ_crop": 47.745845725613485, "PSNR_loc": 21.181631575850673, "SSIM_loc": 0.843030764793486, "LPIPS_loc": 0.17594784468030775, "DreamSim_loc": 0.060776721627595835, "n_clips": 157}
|