@@ -343,7 +343,7 @@ <h2 class="title is-4">
343343 </ h2 >
344344 </ div >
345345 < div style ="text-align: center; ">
346- < img id ="teaser " width ="85 % " src ="static/images/teaser.png ">
346+ < img id ="teaser " width ="80 % " src ="static/images/teaser.png ">
347347 </ div >
348348
349349< section class ="section ">
@@ -373,7 +373,12 @@ <h2 class="title is-4">
373373 We test VideoScore2 on the test set of VideoFeedback2, containing 500 videos with human scores from three dimensions.
374374 Below we show the results of some scoring/reward models for image or video, like VideoReward, VisionReward, Q-Insight, DeQA-Score, and MLLM-prompting methods like Claude-Sonnet-4, GPT-5, Gemini-2.5-Pro, etc and our VideoScore2.
375375 </ p >
376- < img id ="res_vs2_bench " width ="70% " src ="static/images/res_vs2_bench.png ">
376+ < centering >
377+ < div style ="text-align: center; ">
378+ < img id ="res_vs2_bench " width ="70% " src ="static/images/res_vs2_bench.png ">
379+ </ div >
380+ </ centering >
381+
377382 </ div >
378383 </ div >
379384 </ div >
@@ -394,7 +399,11 @@ <h2 class="title is-4">
394399Preference benchmark results include ties. As shown in Table below, while VideoScore2 is
395400 not always the top model on each benchmark, it achieves the highest overall average.
396401 </ p >
397- < img id ="res_ood_bench " width ="70% " src ="static/images/res_ood_bench.png ">
402+ < centering >
403+ < div style ="text-align: center; ">
404+ < img id ="res_ood_bench " width ="70% " src ="static/images/res_ood_bench.png ">
405+ </ div >
406+ </ centering >
398407 </ div >
399408 </ div >
400409 </ div >
@@ -415,7 +424,11 @@ <h2 class="title is-4">
415424 Comparison on VBench shows BoN consistently outperforms random sampling, confirm
416425ing that VIDEOSCORE2 effectively guides higher-quality selection.
417426 </ p >
418- < img id ="BoN " width ="70% " src ="static/images/BoN.png ">
427+ < centering >
428+ < div style ="text-align: center; ">
429+ < img id ="BoN " width ="70% " src ="static/images/BoN.png ">
430+ </ div >
431+ </ centering >
419432 </ div >
420433 </ div >
421434 </ div >
@@ -432,11 +445,13 @@ <h2 class="title is-3">
432445 </ h2 >
433446 </ div >
434447</ section >
435-
436- < img id ="cs1 " width ="70% " src ="static/images/case1.png ">
437- < img id ="cs2 " width ="70% " src ="static/images/case2.png ">
438- < img id ="cs3 " width ="70% " src ="static/images/case3.png ">
439-
448+ < centering >
449+ < div style ="text-align: center; ">
450+ < img id ="cs1 " width ="50% " src ="static/images/case1.png ">
451+ < img id ="cs2 " width ="50% " src ="static/images/case2.png ">
452+ < img id ="cs3 " width ="50% " src ="static/images/case3.png ">
453+ </ div >
454+ </ centering >
440455
441456< section class ="section " id ="BibTeX ">
442457 < div class ="container is-max-desktop content ">
0 commit comments