Spaces:
Running
Running
File size: 5,764 Bytes
565f995 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 |
import streamlit as st
from app.draw_diagram import *
def dashboard():
with st.container():
st.title("AudioBench")
st.markdown("""
[gh]: https://github.com/AudioLLMs/AudioBench
[][gh]
[][gh]
""")
audio_url = "https://arxiv.org/abs/2406.16020"
st.divider()
st.markdown("#### [AudioBench](%s)" % audio_url)
st.markdown("##### :dizzy: A comprehensive evaluation benchmark designed for general instruction-following audiolanguage models")
st.markdown('''
''')
with st.container():
left_co, center_co, right_co = st.columns([0.5,1, 0.5])
with center_co:
st.image("./style/audio_overview.png",
caption="Overview of the datasets in AudioBench.",
use_column_width = True)
st.markdown('''
''')
st.markdown("###### :dart: Our Benchmark includes: ")
cols = st.columns(10)
cols[1].metric(label="Tasks", value="8", delta="Tasks", delta_color="off")
cols[2].metric(label="Datasets", value="26", delta="Datasets", delta_color="off")
cols[3].metric(label="Test On", value="4", delta="Models", delta_color="off")
# st.markdown("###### :dart: Supported Models and Datasets: ")
# sup = pd.DataFrame(
# {"Dataset": "LibriSpeech-Clean",
# "Category": st.selectbox('category', ['Speech Understanding']),
# "Task": st.selectbox('task', ['Automatic Speech Recognition']),
# "Metrics": st.selectbox('metrics', ['WER']),
# "Status":True}
# )
# st.data_editor(sup, num_rows="dynamic")
st.divider()
with st.container():
st.markdown("##### Citations")
st.markdown('''
:round_pushpin: AudioBench Paper \n
@article{wang2024audiobench,
title={AudioBench: A Universal Benchmark for Audio Large Language Models},
author={Wang, Bin and Zou, Xunlong and Lin, Geyu and Sun, Shuo and Liu, Zhuohan and Zhang, Wenyu and Liu, Zhengyuan and Aw, AiTi and Chen, Nancy F},
journal={arXiv preprint arXiv:2406.16020},
year={2024}
}
''')
def speech_understanding():
st.title("Speech Understanding")
filters_levelone = ['ASR', 'SQA', 'SI']
sort_leveltwo = []
left, center, _, middle,right = st.columns([0.2, 0.2, 0.2, 0.2 ,0.2])
with left:
filter_1 = st.selectbox('Select Category', filters_levelone)
with middle:
if filter_1 == filters_levelone[0]:
sort_leveltwo = ['LibriSpeech-Test-Clean', 'LibriSpeech-Test-Other', 'Common-Voice-15-En-Test', 'Peoples-Speech-Test',
'GigaSpeech-Test', 'Tedlium3-Test','Tedlium3-Longform-Test', 'Earning-21-Test', 'Earning-22-Test']
elif filter_1 == filters_levelone[1]:
sort_leveltwo = ['CN-College-Listen-Test', 'SLUE-P2-SQA5-Test', 'DREAM-TTS-Test', 'Public-SG-SpeechQA-Test']
elif filter_1 == filters_levelone[2]:
sort_leveltwo = ['OpenHermes-Audio-Test', 'ALPACA-Audio-Test']
sort = st.selectbox("Sort Dataset", sort_leveltwo)
with right:
sorted = st.selectbox('by', ['Ascending', 'Descending'])
if filter_1 or sort or sorted:
draw('su',filter_1, sort, sorted)
else:
draw('su', 'ASR', 'LibriSpeech-Test-Clean', 'Descending')
def audio_scene_understanding():
st.title("Audio Scence Understanding")
filters_levelone = ['AQA', 'AC']
sort_leveltwo = []
left, center, _, middle,right = st.columns([0.2, 0.2, 0.2, 0.2 ,0.2])
with left:
filter_1 = st.selectbox('Select Category', filters_levelone)
with middle:
if filter_1 == filters_levelone[0]:
sort_leveltwo = ['Clotho-AQA-Test', 'WavCaps-QA-Test', 'AudioCaps-QA-Test']
elif filter_1 == filters_levelone[1]:
sort_leveltwo = ['WavCaps-Test', 'AudioCaps-Test']
sort = st.selectbox("Sort Dataset", sort_leveltwo)
with right:
sorted = st.selectbox('by', ['Ascending', 'Descending'])
if filter_1 or sort or sorted:
draw('asu',filter_1, sort, sorted)
else:
draw('asu', 'AQA', 'Clotho-AQA-Test', 'Descending')
def voice_understanding():
st.title("Voice Understanding")
filters_levelone = ['ER', 'AR', 'GR']
sort_leveltwo = []
left, center, _, middle,right = st.columns([0.2, 0.2, 0.2, 0.2 ,0.2])
with left:
filter_1 = st.selectbox('Select Category', filters_levelone)
with middle:
if filter_1 == filters_levelone[0]:
sort_leveltwo = ['IEMOCAP-Emotion-Test', 'MELD-Sentiment-Test', 'MELD-Emotion-Test']
elif filter_1 == filters_levelone[1]:
sort_leveltwo = ['VoxCeleb1-Accent-Test']
elif filter_1 == filters_levelone[2]:
sort_leveltwo = ['VoxCeleb1-Gender-Test', 'IEMOCAP-Gender-Test']
sort = st.selectbox("Sort Dataset", sort_leveltwo)
with right:
sorted = st.selectbox('by', ['Ascending', 'Descending'])
if filter_1 or sort or sorted:
draw('vu',filter_1, sort, sorted)
else:
draw('vu', 'ER', 'IEMOCAP-Emotion-Test', 'Descending')
|