IndexTTS2：情感表达和持续时间控制的自回归零镜头文本转语音的突破
IndexTTS2：情感表达和持续时间控制的自回归零镜头文本转语音的突破
IndexTTS2：情感表达和持续时间控制的自回归零镜头文本转语音的突破 IndexTTS2：情感表达和持续时间控制的自回归零镜头文本转语音的突破 Modified July 16, 2025 Audio or video is not supported Please download the file and open it with the appropriate software. Let the Bullets Fly.mp4 · 4.44MB Let the Bullets Fly 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Empresses in the Palace.mp4 · 15.13MB Empresses in the Palace 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Empresses in the Palace 2.mp4 · 16.19MB Empresses in the Palace 2 00:00 prompt Audio or video is not supported gt Audio or video is not supported indextts2 943 gt0.75 Audio or video is not supported indextts2 943 gt Audio or video is not supported indextts2 943 gt1.25 Audio or video is not supported maskgct 943 gt0.75 Audio or video is not supported maskgct 943 gt Audio or video is not supported maskgct 943 gt1.25 Audio or video is not supported f5 943 gt0.75 Audio or video is not supported f5 943 gt Audio or video is not supported f5 943 gt1.25 Audio or video is not supported prompt (1) Audio or video is not supported gt (1) Audio or video is not supported indextts2 1037 gt0.75 Audio or video is not supported indextts2 1037 gt Audio or video is not supported indextts2 1037 gt1.25 Audio or video is not supported maskgct 1037 gt0.75 Audio or video is not supported maskgct 1037 gt Audio or video is not supported maskgct 1037 gt1.25 Audio or video is not supported f5 1037 gt0.75 Audio or video is not supported f5 1037 gt Audio or video is not supported f5 1037 gt1.25 Audio or video is not supported prompt (2) Audio or video is not supported gt (2) Audio or video is not supported indextts2 289 gt0.75 Audio or video is not supported indextts2 289 gt Audio or video is not supported indextts2 289 gt1.25 Audio or video is not supported maskgct 289 gt0.75 Audio or video is not supported maskgct 289 gt Audio or video is not supported maskgct 289 gt1.25 Audio or video is not supported f5 289 gt0.75 Audio or video is not supported f5 289 gt Audio or video is not supported f5 289 gt1.25 Audio or video is not supported prompt (3) Audio or video is not supported gt (3) Audio or video is not supported indextts2 319 gt0.75 Audio or video is not supported indextts2 319 gt Audio or video is not supported indextts2 319 gt1.25 Audio or video is not supported maskgct 319 gt0.75 Audio or video is not supported maskgct 319 gt Audio or video is not supported maskgct 319 gt1.25 Audio or video is not supported f5 319 gt0.75 Audio or video is not supported f5 319 gt Audio or video is not supported f5 319 gt1.25 Audio or video is not supported 08 angry 3 Audio or video is not supported 600 Audio or video is not supported 600 (1) Audio or video is not supported 600 (2) Audio or video is not supported 600 (3) Audio or video is not supported 600 (4) Audio or video is not supported 12 angry 2 Audio or video is not supported 929 Audio or video is not supported 929 (1) Audio or video is not supported 929 (2) Audio or video is not supported 929 (3) Audio or video is not supported 929 (4) Audio or video is not supported 10 cry 2 Audio or video is not supported 788 Audio or video is not supported 788 (1) Audio or video is not supported 788 (2) Audio or video is not supported 788 (3) Audio or video is not supported 788 (4) Audio or video is not supported 05 cry 1 Audio or video is not supported 361 Audio or video is not supported 361 (1) Audio or video is not supported 361 (2) Audio or video is not supported 361 (3) Audio or video is not supported 361 (4) Audio or video is not supported 11 fear 2 Audio or video is not supported 881 Audio or video is not supported 881 (1) Audio or video is not supported 881 (2) Audio or video is not supported 881 (3) Audio or video is not supported 881 (4) Audio or video is not supported 08 fear 3 Audio or video is not supported 635 Audio or video is not supported 635 (1) Audio or video is not supported 635 (2) Audio or video is not supported 635 (3) Audio or video is not supported 635 (4) Audio or video is not supported 03 low 3 Audio or video is not supported 239 Audio or video is not supported 239 (1) Audio or video is not supported 239 (2) Audio or video is not supported 239 (3) Audio or video is not supported 239 (4) Audio or video is not supported 12 low 3 Audio or video is not supported 993 Audio or video is not supported 993 (1) Audio or video is not supported 993 (2) Audio or video is not supported 993 (3) Audio or video is not supported 993 (4) Audio or video is not supported 06 calm 2 Audio or video is not supported 439 Audio or video is not supported 439 (1) Audio or video is not supported 439 (2) Audio or video is not supported 439 (3) Audio or video is not supported 439 (4) Audio or video is not supported bage Unable to preview. Please download the file. 10 fear 2 Unable to preview. Please download the file. bage 01 0.0 Unable to preview. Please download the file. bage 01 0.6 Unable to preview. Please download the file. bage 01 1.0 Unable to preview. Please download the file. bage 01 1.4 Unable to preview. Please download the file. 08 angry 3 (1) Unable to preview. Please download the file. bage 03 0.0 Unable to preview. Please download the file. bage 03 0.6 Unable to preview. Please download the file. bage 03 1.0 Unable to preview. Please download the file. bage 03 1.4 Unable to preview. Please download the file. jieshuonan Unable to preview. Please download the file. 12 low 3 (1) Unable to preview. Please download the file. jieshuonan 16 0.0 Unable to preview. Please download the file. jieshuonan 16 0.6 Unable to preview. Please download the file. jieshuonan 16 1.0 Unable to preview. Please download the file. jieshuonan 16 1.4 Unable to preview. Please download the file. 07 surprise 3 Unable to preview. Please download the file. jieshuonan 25 0.0 Unable to preview. Please download the file. jieshuonan 25 0.6 Unable to preview. Please download the file. jieshuonan 25 1.0 Unable to preview. Please download the file. jieshuonan 25 1.4 Unable to preview. Please download the file. xindong Unable to preview. Please download the file. 12 low 3 (2) Unable to preview. Please download the file. 有些人走了就再也没有回来过，所以等待和犹豫是这个世界上最无情的杀手。 xindong 17 0.0 Unable to preview. Please download the file. xindong 17 0.6 Unable to preview. Please download the file. xindong 17 1.0 Unable to preview. Please download the file. xindong 17 1.4 Unable to preview. Please download the file. 04 happy 3 Unable to preview. Please download the file. 做一个温暖的人，将岁月里的凝重、安暖，写意成简单，将过往的风景，安放在清浅的时光中。 xindong 21 0.0 Unable to preview. Please download the file. xindong 21 0.6 Unable to preview. Please download the file. v xindong 21 1.0 Unable to preview. Please download the file. xindong 21 1.4 Unable to preview. Please download the file. 2.3 使用文本描述 在这种替代方法中，我们将情感音频参考替换为描述性文本提示，同时保持音色参考音频不变。这允许模型仅根据基于文本的情感线索而不是音频信号来调节情感表达。这样的设置有助于研究模型生成由语言描述指导的情感表达语音的能力，从而在不依赖情感音频提示的情况下深入了解文本驱动的情绪控制的有效性。 音色 音频 提示 情感描述 文本 音频 bage (1) Unable to preview. Please download the file. 我感到非常沮丧 这究竟是我的福，还是我的孽？岂止是皇上错了，我更是错了！这几年的情爱与时光，究竟是错付了！ bage gen 1 Unable to preview. Please download the file. 有点快乐，哈哈 bage gen 2 Unable to preview. Please download the file. 巨巨巨巨巨巨巨巨难过 bage gen 3 Unable to preview. Please download the file. 我感到非常沮丧 这是我的祝福，还是我的诅咒？不仅仅是皇帝错了——我错得更厉害！这些年的爱和奉献......到头来，他们只不过是一颗浪费的心吗？ bage gen 11 Unable to preview. Please download the file. 有点快乐，哈哈 bage gen 12 Unable to preview. Please download the file. 巨巨巨巨巨巨巨巨难过 bage gen 13 Unable to preview. Please download the file. tangxi Unable to preview. Please download the file. 书桓走的第一天，想他，想他，想他。 书桓走的第一天，想他，想他，想他。 tangxi gen 1 Unable to preview. Please download the file. 书桓离开的第一天，我所做的就是想念他。想念他。想念他。 tangxi gen 11 Unable to preview. Please download the file. 书桓走的第二天，想他，想他，想他。 书桓走的第二天，想他，想他，想他。 tangxi gen 2 Unable to preview. Please download the file. 第二天，书桓走了，但我仍然想念他。我想念他。我想念他。 tangxi gen 12 Unable to preview. Please download the file. 书桓走了第三天了，想他，想他，想他，发疯一样的想他。 书桓走了第三天了，想他，想他，想他，发疯一样的想他。 tangxi gen 3 Unable to preview. Please download the file. 书桓走了第三天......我仍然想念他。想念他。想念他。我想念他，就像我失去了理智一样。 tangxi gen 13 Unable to preview. Please download the file. qingsong Unable to preview. Please download the file. 超级无敌爆炸angry的情感，就像刚中了彩票被人偷拿了 你问他为什么我没谈恋爱，我就失恋了，你问他，为什么这么对我，我以为我会问，可是我见到他之后，我就不想问了，因为人家根本就不想说，人家甚至都不想见到你，我为什么在那儿犯贱呢，所以我不是放过他，我是想放过我自己，人家不联系你怎么了，不回你微信怎么了，伤害你又怎么了，你算他谁啊? qingsong 1 Unable to preview. Please download the file. tangxi (1) Unable to preview. Please download the file. 又生气又委屈 我为什么非得知道发生了什么呢，我不就是想给自己找一个原因嘛，我只是想找一个原因，我原谅他，可是我为什么非得原谅他呢，我干嘛把自己搞得这么卑微啊? tangxi 3 Unable to preview. Please download the file. bage (2) Unable to preview. Please download the file. 我们正在做一些神奇的事情，给我来一种又fear，但是又有点开心的情感。 这游戏太刺激了，心跳都快停了...但我们又能感受到那种挑战未知的兴奋和快乐。说真的，我现在是又害怕又期待，紧张得手心都在冒汗！ bage 15 Unable to preview. Please download the file. 3 零样本 TTS 与其他 TTS 的比较（开源数据） 3.1 单语语音合成 音频提示 文本 GroundTruth 公司 指数 TTS2 指数 TTS 掩码 GCT CosyVoice2 0296 Unable to preview. Please download the file. 家居养娃的李娜又重新出现在媒体大众的面前 0296 (1) Unable to preview. Please download the file. 0296 (2) Unable to preview. Please download the file. 0296 (3) Unable to preview. Please download the file. 0296 (4) Unable to preview. Please download the file. 0296 (5) Unable to preview. Please download the file. 0735 Unable to preview. Please download the file. 这是南极洲仅有的三个已知拥有恐龙化石的地层中的两个。 0735 (1) Unable to preview. Please download the file. 0735 (2) Unable to preview. Please download the file. 0735 (3) Unable to preview. Please download the file. 0735 (4) Unable to preview. Please download the file. 0735 (5) Unable to preview. Please download the file. 1013 Unable to preview. Please download the file. 男人看着他，没有回应。 1013 (1) Unable to preview. Please download the file. 1013 (2) Unable to preview. Please download the file. 1013 (3) Unable to preview. Please download the file. 1013 (4) Unable to preview. Please download the file. 1013 (5) Unable to preview. Please download the file. 0829 Unable to preview. Please download the file. 胡萝卜凉拌或炒鸡蛋味道都是棒极的，胡萝卜骄傲地说。 0829 (1) Unable to preview. Please download the file. 0829 (2) Unable to preview. Please download the file. 0829 (3) Unable to preview. Please download the file. 0829 (4) Unable to preview. Please download the file. 0829 (5) Unable to preview. Please download the file. 1806 Unable to preview. Please download the file. 鲁道夫毫无阻碍地来到了自己的房子，而利奥卡迪亚的父母则伤心欲绝 1806 (1) Unable to preview. Please download the file. 1806 (2) Unable to preview. Please download the file. 1806 (3) Unable to preview. Please download the file. 1806 (4) Unable to preview. Please download the file. 1806 (5) Unable to preview. Please download the file. 1358 Unable to preview. Please download the file. 那些袖珍衣服挂在架子上，远远看上去就像一幅画，可漂亮了。 1358 (1) Unable to preview. Please download the file. 1358 (2) Unable to preview. Please download the file. 1358 (3) Unable to preview. Please download the file. 1358 (4) Unable to preview. Please download the file. 1358 (5) Unable to preview. Please download the file. 原文链接：https://index tts.github.io/index tts2.github.io/ Audio or video is not supported Please download the file and open it with the appropriate software. Let the Bullets Fly.mp4 · 4.44MB Let the Bullets Fly 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Empresses in the Palace.mp4 · 15.13MB Empresses in the Palace 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Let the Bullets Fly.mp4 · 4.44MB Let the Bullets Fly 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Let the Bullets Fly.mp4 · 4.44MB Let the Bullets Fly 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Let the Bullets Fly.mp4 · 4.44MB Let the Bullets Fly 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Empre