DMD icon indicating copy to clipboard operation
DMD copied to clipboard

An official implementation of "Decoupled Multimodal Distilling for Emotion Recognition" in PyTorch. (CVPR 2023 highlight)

Results 12 DMD issues
Sort by recently updated
recently updated
newest added

Dear author, congratulations on your outstanding achievements. Without modifying the code, I obtained the following results on the MOSI dataset: ACC2 84.4% and F1 84.4%, which are slightly lower than...

Impressive idea! But i cannot reproduce the performance reported in the original paper using the default settings and parameters, particularly with respect to the F1 and ACC2 metrics (with a...

在mosi数据集上的训练集准确率比验证集和测试集高很多,请问是参数没有调至合适的还是模型较复杂,数据量不够还是训练集数据泄露导致的结果

![image](https://github.com/user-attachments/assets/14cbfd3e-45ee-4447-804d-29917bfd2887) 这里只体现了相同类别和不相同类别的特征学习啊

Hello, may I ask if the accuracy result in this paper is a good thing of has0 and no-has0? What's the difference between aligned and misaligned results? What method is...

requests.exceptions.ProxyError: (MaxRetryError("HTTPSConnectionPool(host='huggingface.co', port=443): Max retries exceeded with url: /bert-base-uncased/resolve/main/tokenizer_config.json (Caused by ProxyError('Unable to connect to proxy', SSLError(SSLZeroReturnError(6, 'TLS/SSL connection has been closed (EOF) (_ssl.c:1135)'))))"), '(Request ID: 737b9db1-9199-430a-ac90-7c9956c977c0)')

As I know, **cos(X, Y)=0** means X and Y are orthogonal, meanwhile the minimum value of **cos(X, Y)=-1** means X and Y are opposite. So, as written in your manuscript,...

It is a nice job. And I have some question about the gd_prior used in th class get_distillation_kernel in code. What are the principles for setting up this param? And...