Self-MM
Self-MM copied to clipboard
Codes for paper "Learning Modality-Specific Representations with Self-Supervised Multi-Task Learning for Multimodal Sentiment Analysis"
too many values to unpack (expected 7)
您好!我想请问下论文上面写的实验结果是实验代码跑出来写在表格里面的结果吗?是5个不同随机种子跑出来结果的均值,还是? 目前,我跑出来的实验结果(即写在表格里面的)和论文上面的结果不符合
作者您好,读了您这篇文章后,我深受启发!但我在复现模型到自己任务后,发现GPU上运算耗时只占小部分,大部分计算时间是在CPU上,进行如list append操作及三个update函数上,导致CPU占用率很高,GPU利用率却很低。  我想请教您训练模型的时候,资源消耗也是类似的情况吗?还是我的代码实现可能有问题呢?期待您的回复,谢谢!
视频音频特征提取
我刚刚读到了您在 AAAI 2021 上发表的精彩论文,题为“Learning Modality-Specific Representations with Self-Supervised Multi-Task Learning for Multimodal Sentiment Analysis”。我想请问一下您,在论文中您并未阐述视频音频特征提取具体使用的是什么,您方便告知一下吗?视频模块使用的是MediaPipe还是OpenFace?音频模块使用的是Librosa、OpenSMILE还是Wav2Vec呢? 抱歉耽误您的时间了,期待您的回复!谢谢!
作者您好
Hello, I am a beginner, I would like to ask how to reproduce Figure 3 in the paper?
数据集我选择了原始的SIMS数据集,使用DataPre.py文件对其进行了特征提取,最终得到了featurs.pkl。然后我将其带入了self模型进行训练,它在“/Self-MM-main/data/load_data.py”出现了报错,报错如下图:  我发现通过DataPre提取的features.pkl文件,没有包含regression_labels_T、regression_labels_V、regression_labels_A、classification_labels_A、classification_labels_T、classification_labels_V的值。而您在百度网盘中提供的SIMS/Processed/unaligned_39.pkl中是包含这些值的。 请问我是不是在特征提取这一步中做错了呢?
Hi, could you please give some final Self-MM checkpoints on different datasets such as MOSI and MOSEI? Thx.
Hello Big Man, I see in today's pycharm, like audio_ x.size(1) function method is no longer valid, and my question is whether it returns max_ Len or actual length? thx...
Bumps [numpy](https://github.com/numpy/numpy) from 1.21.0 to 1.22.0. Release notes Sourced from numpy's releases. v1.22.0 NumPy 1.22.0 Release Notes NumPy 1.22.0 is a big release featuring the work of 153 contributors spread...