search

thuiar / MMSA

MMSA is a unified framework for Multimodal Sentiment Analysis.
MIT License
634 stars 104 forks source link

想获取一下SIMSv2这个数据集的特征提取的配置 #73

Open tangyc314 opened 10 months ago

tangyc314 commented 10 months ago

"audio": { "tool": "opensmile", "sample_rate": 16000, "args": { "feature_set": "eGeMAPSv02", "feature_level": "LowLevelDescriptors", "start": null, "end": null } }, "video": { "tool": "openface", "fps": 25, "multiFace": { "enable": false, "device": "cuda:0", "facedetScale": 0.25, "minTrack": 10, "numFailedDet": 10, "minFaceSize": 1, "cropScale": 0.4 }, "average_over": 1, "args": { "hogalign": false, "simalign": false, "nobadaligned": false, "landmark_2D": true, "landmark_3D": false, "pdmparams": false, "head_pose": false, "action_units": true, "gaze": false, "tracked": false } 以以上参数提取特征，获得的vision特征向量范围不是-1到1，而是会包含一些很大的数据如5.18e+02，是有哪里需要调整吗

cherishPre commented 8 months ago

您好，使用MMSA-FET中openface工具获取到的特征向量具有较大数值属于正常情况，可进一步使用归一化将特征向量范围限制在-1到1之间。