Skip to content

Commit 4518a0e

Browse files
sword9-keepersword9-keeper
authored
Merge pull request #16 from byrTony-Frankzyq/master
fix typo, rearrange the sections
2 parents 42bd29b + 2b003dd commit 4518a0e

1 file changed

Lines changed: 67 additions & 67 deletions

File tree

‎funasr/index.html‎

Lines changed: 67 additions & 67 deletions
Original file line numberDiff line numberDiff line change
@@ -340,73 +340,6 @@ <h2 class="title is-3 has-text-centered">
340340
<p class="lang-zh">点击播放试听 Fun-ASR 在不同场景下的识别效果。</p>
341341
</div>
342342

343-
<!-- ========================================== -->
344-
<!-- 板块：古诗词识别 -->
345-
<!-- ========================================== -->
346-
<div class="asr-section">
347-
<h3 class="asr-category-title">
348-
<span class="lang-en">Classical Chinese Poetry Recognition</span>
349-
<span class="lang-zh">古诗词识别专项优化</span>
350-
</h3>
351-
<p class="lang-en">Fun-ASR v1.5 introduces dedicated optimization for classical Chinese poetry recitation. Classical poetry poses unique challenges: concise classical grammar, strict rhyming and rhythmic patterns, frequent use of典故 (allusions),异体字 (variant characters), and古今异义词 (words with shifted meanings), as well as non-natural speech features like拖腔 (elongated tones), pauses, and吟咏 (chanting). In internal evaluation, v1.5 achieves a character-level accuracy of 97% on classical poetry.</p>
352-
<p class="lang-zh">Fun-ASR v1.5 对中文古诗词识别进行了专项优化。不同于现代口语，古诗词具有文言语法简练、押韵严格节奏固定、多用典故异体字古今异义词、诵读时存在拖腔停顿吟咏等非自然语流特征等挑战。在内部评测集中，v1.5 对古诗词的字符级准确率达到 97%。</p>
353-
<div class="columns is-multiline">
354-
355-
<!-- 蓬山此去无多路 -->
356-
<div class="column is-12">
357-
<div class="sample-card">
358-
<div class="columns is-vcentered">
359-
<div class="column is-3 has-text-centered">
360-
<span class="sample-tag" style="margin-bottom: 0.5rem; display: inline-block;">
361-
<span class="lang-en">Classical Poetry</span>
362-
<span class="lang-zh">古诗词</span>
363-
</span>
364-
<audio controls style="width: 100%;">
365-
<source src="static/audios/poetry_pengshan_v15.wav" type="audio/wav">
366-
</audio>
367-
</div>
368-
<div class="column is-9">
369-
<div class="transcript-box">
370-
<span class="transcript-label">
371-
<span class="lang-en">ASR Output</span>
372-
<span class="lang-zh">识别结果</span>
373-
</span>
374-
<p class="transcript-text">蓬山此去无多路，青鸟殷勤为探看。</p>
375-
</div>
376-
</div>
377-
</div>
378-
</div>
379-
</div>
380-
381-
<!-- 子夏曰 -->
382-
<div class="column is-12">
383-
<div class="sample-card">
384-
<div class="columns is-vcentered">
385-
<div class="column is-3 has-text-centered">
386-
<span class="sample-tag" style="margin-bottom: 0.5rem; display: inline-block;">
387-
<span class="lang-en">Classical Text</span>
388-
<span class="lang-zh">古文</span>
389-
</span>
390-
<audio controls style="width: 100%;">
391-
<source src="static/audios/poetry_zixia_v15.wav" type="audio/wav">
392-
</audio>
393-
</div>
394-
<div class="column is-9">
395-
<div class="transcript-box">
396-
<span class="transcript-label">
397-
<span class="lang-en">ASR Output</span>
398-
<span class="lang-zh">识别结果</span>
399-
</span>
400-
<p class="transcript-text">子夏曰，博学而笃志，切问而近思，仁在其中矣。</p>
401-
</div>
402-
</div>
403-
</div>
404-
</div>
405-
</div>
406-
407-
</div>
408-
</div>
409-
410343
<!-- ========================================== -->
411344
<!-- 板块：中文方言识别 -->
412345
<!-- ========================================== -->
@@ -771,6 +704,73 @@ <h3 class="asr-category-title">
771704

772705
</div>
773706
</div>
707+
<!-- ========================================== -->
708+
<!-- 板块：古诗词识别 -->
709+
<!-- ========================================== -->
710+
<div class="asr-section">
711+
<h3 class="asr-category-title">
712+
<span class="lang-en">Classical Chinese Poetry Recognition</span>
713+
<span class="lang-zh">古诗词识别专项优化</span>
714+
</h3>
715+
<p class="lang-en">Fun-ASR v1.5 introduces dedicated optimization for classical Chinese poetry recitation. Unlike modern spoken language, classical poetry poses unique challenges: concise classical grammar with omitted subjects, predicates, and objects; strict rhyming and fixed rhythmic patterns (e.g., five-character and seven-character regulated verse); frequent use of allusions, variant characters, and words with shifted meanings across eras; and non-natural speech features such as elongated tones, deliberate pauses, and chanting rhythms. In internal evaluation, v1.5 achieves a character-level accuracy of 97% on classical poetry.</p>
716+
<p class="lang-zh">Fun-ASR v1.5 对中文古诗词识别进行了专项优化。不同于现代口语，古诗词具有文言语法简练、押韵严格节奏固定、多用典故异体字古今异义词、诵读时存在拖腔停顿吟咏等非自然语流特征等挑战。在内部评测集中，v1.5 对古诗词的字符级准确率达到 97%。</p>
717+
<div class="columns is-multiline">
718+
719+
<!-- 蓬山此去无多路 -->
720+
<div class="column is-12">
721+
<div class="sample-card">
722+
<div class="columns is-vcentered">
723+
<div class="column is-3 has-text-centered">
724+
<span class="sample-tag" style="margin-bottom: 0.5rem; display: inline-block;">
725+
<span class="lang-en">Classical Poetry</span>
726+
<span class="lang-zh">古诗词</span>
727+
</span>
728+
<audio controls style="width: 100%;">
729+
<source src="static/audios/poetry_pengshan_v15.wav" type="audio/wav">
730+
</audio>
731+
</div>
732+
<div class="column is-9">
733+
<div class="transcript-box">
734+
<span class="transcript-label">
735+
<span class="lang-en">ASR Output</span>
736+
<span class="lang-zh">识别结果</span>
737+
</span>
738+
<p class="transcript-text">蓬山此去无多路，青鸟殷勤为探看。</p>
739+
</div>
740+
</div>
741+
</div>
742+
</div>
743+
</div>
744+
745+
<!-- 子夏曰 -->
746+
<div class="column is-12">
747+
<div class="sample-card">
748+
<div class="columns is-vcentered">
749+
<div class="column is-3 has-text-centered">
750+
<span class="sample-tag" style="margin-bottom: 0.5rem; display: inline-block;">
751+
<span class="lang-en">Classical Text</span>
752+
<span class="lang-zh">古文</span>
753+
</span>
754+
<audio controls style="width: 100%;">
755+
<source src="static/audios/poetry_zixia_v15.wav" type="audio/wav">
756+
</audio>
757+
</div>
758+
<div class="column is-9">
759+
<div class="transcript-box">
760+
<span class="transcript-label">
761+
<span class="lang-en">ASR Output</span>
762+
<span class="lang-zh">识别结果</span>
763+
</span>
764+
<p class="transcript-text">子夏曰，博学而笃志，切问而近思，仁在其中矣。</p>
765+
</div>
766+
</div>
767+
</div>
768+
</div>
769+
</div>
770+
771+
</div>
772+
</div>
773+
774774
<!-- ========================================== -->
775775
<!-- 板块：标点预测与ITN -->
776776
<!-- ========================================== -->

0 commit comments

Comments
 (0)