CMU-MultimodalSDK
CMU-MultimodalSDK copied to clipboard
A2Zadeh
CMU MultimodalSDK is a machine learning platform for development of advanced multimodal models as well as easily accessing and processing multimodal datasets.
Hi, I need to use the raw data to extract the features myself but I can't find the MOSI data labels. It was suggested before to see this link http://immortal.multicomp.cs.cmu.edu/cache/...
I have trained Multimodal-Infomax model on CMU MOSI dataset. But I do not know how to predict a piece of my self data(a video with audio). Can anyone help me?
platform：windows10 For example: In file:"download_ops.py" : readURL("http://immortal.multicomp.cs.cmu.edu/CMUMOSEI/acoustic/CMU_MOSEI_COVAREP.csd","./hi.csd") the separator is '/' . however, "def __initialize_from_csd(self,resource,destination,validate)" in the file of "computational_sequence.py" destination=os.path.join(destination,resource.split(os.sep)[-1]). resource.split(os.sep)[-1] == resource.split('\')[-1] == "http://immortal.multicomp.cs.cmu.edu/CMUMOSEI/acoustic/CMU_MOSEI_COVAREP.csd".split('\')[-1] != "CMU_MOSEI_COVAREP.csd". If you...
Hello, The description of CMU-MOSEI dataset shows that the total number of speakers is 1000 and the number of topics is 250. Can we know the video files belonging to...
Hello! I saw that some videos in the CMI_MOSEI segmented section are corrupted. The videos only have audio (no sound) and open-cv fails to open the video and extract frames....
Hello everyone, I would like to know which tool you guys used for the process of annotating the data. Looking forward for help :)
By working on this repo and using the up-to-date computational sequences for all datasets, I found that the current versions of MOSEI and MOSI, as found in http://immortal.multicomp.cs.cmu.edu/ and even...
Is there any tutorials on how to release my dataset? And what is the format of computational sequences of MOSI and MOSEI
Hi, Is it possible to align an externally downloaded dataset (eg: booktube dataset) using this SDK? Thanks :)
Hello! I'm basically extracting my own features set from the raw data on CMU-MOSEI and I extract the sentiment/emotions labels from the SDK (using this [tutorial](https://github.com/Justin1904/CMU-MultimodalSDK-Tutorials) ) Then I get...
