@@ -220,6 +220,18 @@ <h1 class="title is-1 publication-title">
220220 </ a >
221221 </ span >
222222
223+ <!-- Paper -->
224+ < span class ="link-block ">
225+ < a href ="https://arxiv.org/abs/2509.12508 " target ="_blank "
226+ class ="external-link button is-normal is-rounded is-dark ">
227+ < span class ="icon ">
228+ < i class ="fas fa-file-pdf "> </ i >
229+ </ span >
230+ < span > Paper</ span >
231+ </ a >
232+ </ span >
233+
234+
223235 <!-- ModelScope Demo -->
224236 < span class ="link-block ">
225237 < a href ="https://modelscope.cn/studios/FunAudioLLM/Fun-ASR-Nano/ " target ="_blank "
@@ -1485,6 +1497,7 @@ <h4 class="table-title">Industry Dataset Performance (WER %)</h4>
14851497 </ div >
14861498
14871499 <!-- 流式识别性能 -->
1500+ <!-- 流式识别性能 -->
14881501 < div class ="performance-table-container ">
14891502 < h4 class ="table-title "> Streaming ASR Performance (WER %)</ h4 >
14901503 < div class ="table-wrapper ">
@@ -1493,52 +1506,64 @@ <h4 class="table-title">Streaming ASR Performance (WER %)</h4>
14931506 < tr >
14941507 < th > Test Set</ th >
14951508 < th > Seed-ASR</ th >
1496- < th > Fun-ASR</ th >
14971509 < th > Fun-ASR-nano</ th >
1510+ < th > Fun-ASR</ th >
14981511 </ tr >
14991512 </ thead >
15001513 < tbody >
15011514 < tr >
15021515 < td > Nearfield</ td >
15031516 < td > 8.64</ td >
1504- < td > 7.00 </ td >
1505- < td > 7.97 </ td >
1517+ < td > 8.10 </ td >
1518+ < td > 6.75 </ td >
15061519 </ tr >
15071520 < tr >
1508- < td > Fairfield </ td >
1521+ < td > Farfield </ td >
15091522 < td > 5.51</ td >
1510- < td > 5.33</ td >
1511- < td > 6.92</ td >
1512- </ tr >
1513- < tr >
1514- < td > Home Scenario</ td >
1515- < td > 9.70</ td >
1516- < td > 5.33</ td >
1517- < td > 6.51</ td >
1523+ < td > 6.38</ td >
1524+ < td > 4.72</ td >
15181525 </ tr >
15191526 < tr >
15201527 < td > Complex Background</ td >
15211528 < td > 15.48</ td >
1522- < td > 12.50 </ td >
1523- < td > 14.83 </ td >
1529+ < td > 15.52 </ td >
1530+ < td > 12.49 </ td >
15241531 </ tr >
15251532 < tr >
15261533 < td > English General</ td >
15271534 < td > 18.78</ td >
1528- < td > 14.74 </ td >
1529- < td > 16.70 </ td >
1535+ < td > 16.46 </ td >
1536+ < td > 14.68 </ td >
15301537 </ tr >
15311538 < tr >
1532- < td > OpenSource Test Sets </ td >
1539+ < td > OpenSource</ td >
15331540 < td > 3.80</ td >
1534- < td > 3.60 </ td >
1535- < td > 5.13 </ td >
1541+ < td > 5.06 </ td >
1542+ < td > 4.08 </ td >
15361543 </ tr >
1537- < tr class ="average-row ">
1538- < td > Average</ td >
1539- < td > 10.32</ td >
1540- < td > 8.08</ td >
1541- < td > 9.68</ td >
1544+ < tr >
1545+ < td > Dialect</ td >
1546+ < td > -</ td >
1547+ < td > 30.72</ td >
1548+ < td > 18.25</ td >
1549+ </ tr >
1550+ < tr >
1551+ < td > Accent</ td >
1552+ < td > -</ td >
1553+ < td > 15.42</ td >
1554+ < td > 11.49</ td >
1555+ </ tr >
1556+ < tr >
1557+ < td > Lyrics</ td >
1558+ < td > -</ td >
1559+ < td > 31.54</ td >
1560+ < td > 22.05</ td >
1561+ </ tr >
1562+ < tr >
1563+ < td > Hiphop</ td >
1564+ < td > -</ td >
1565+ < td > 36.55</ td >
1566+ < td > 28.90</ td >
15421567 </ tr >
15431568 </ tbody >
15441569 </ table >
@@ -2106,7 +2131,11 @@ <h2 class="title">
21062131 primaryClass={cs.CL},
21072132 url={https://arxiv.org/abs/2509.12508},
21082133}</ code > </ pre >
2109- </ div >
2134+ <!-- add a note that this paper will be updated sooner, please update your notation if needed -->
2135+ < div style ="margin-top: 1rem; font-size: 0.9rem; color: var(--text-secondary); font-style: italic; ">
2136+ < span class ="lang-en "> note: This paper will be updated soon on arXiv. Please update your citation if needed.</ span >
2137+ < span class ="lang-zh "> 注:该论文将在 arXiv 上更新。 如有需要,请更新您的引用。</ span >
2138+ </ div >
21102139</ section >
21112140<!--End BibTex citation -->
21122141
0 commit comments