Skip to main content

Browse the archive

https://github.com/desmond-ong/cheem-omg-empathy
05 April 2024, 19:47:11 UTC
sort by:
Revision Author Date Message Commit Date
92a6aa9 Xuan 13 December 2018, 03:01:39 UTC Made README filenames consistent. 13 December 2018, 03:01:39 UTC
52e93d0 Xuan 13 December 2018, 02:59:23 UTC Made calculateCCC also calculate std. 13 December 2018, 02:59:23 UTC
6e996e2 Arushi Goel 13 December 2018, 01:47:39 UTC Update README.md 13 December 2018, 01:47:39 UTC
b2d80af Desmond Ong 12 December 2018, 10:16:06 UTC added link to saved model file 12 December 2018, 10:16:06 UTC
b32d900 Xuan 07 December 2018, 11:00:42 UTC Turn subfolder listing into a list in Markdown (fixed). 07 December 2018, 11:00:42 UTC
d45ab25 Xuan 07 December 2018, 11:00:07 UTC Turn subfolder listing into a list in Markdown. 07 December 2018, 11:00:07 UTC
9c15546 Xuan 07 December 2018, 10:59:05 UTC Clean up base directory a bit. 07 December 2018, 10:59:05 UTC
cd75059 Xuan 07 December 2018, 10:50:17 UTC Merge branch 'master' of github.com:desmond-ong/cheem-omg-empathy 07 December 2018, 10:50:17 UTC
c40537d Xuan 07 December 2018, 10:50:06 UTC Updated main README.md 07 December 2018, 10:50:06 UTC
783175f Son Nguyen 07 December 2018, 10:43:29 UTC Update Readme.md 07 December 2018, 10:43:29 UTC
f85039f Xuan 07 December 2018, 10:43:14 UTC Changed README heading for multimodal 07 December 2018, 10:43:14 UTC
01ba837 Xuan 07 December 2018, 10:40:50 UTC Add Markdown and Jupyter Notebook guide to training and running the multi-modal models. 07 December 2018, 10:40:50 UTC
74da2cd Son Nguyen 07 December 2018, 10:27:12 UTC Update Readme.md 07 December 2018, 10:27:12 UTC
64d6a0a Son Nguyen 07 December 2018, 10:26:45 UTC Update Readme.md 07 December 2018, 10:26:45 UTC
911a96e Son Nguyen 07 December 2018, 10:14:52 UTC Update TextOnly Model 07 December 2018, 10:14:52 UTC
39b2d40 Son Nguyen 07 December 2018, 10:14:03 UTC update src 07 December 2018, 10:14:03 UTC
41bc60b Son Nguyen 07 December 2018, 10:04:11 UTC Text only model (G2+P2) 07 December 2018, 10:04:11 UTC
eaf3926 Son Nguyen 07 December 2018, 10:00:39 UTC Text only model (G2+P2) 07 December 2018, 10:00:39 UTC
6bf7fad Son Nguyen 07 December 2018, 09:54:07 UTC Merge branch 'master' of https://github.com/desmond-ong/cheem-omg-empathy 07 December 2018, 09:54:07 UTC
fcf804a Son Nguyen 07 December 2018, 09:52:21 UTC Text only models (G2+P2) 07 December 2018, 09:52:21 UTC
8f41b0f Arushi Goel 07 December 2018, 09:51:44 UTC Update Readme.md 07 December 2018, 09:51:44 UTC
78178cc Arushi Goel 07 December 2018, 09:31:42 UTC Update Readme.md 07 December 2018, 09:31:42 UTC
7c9c117 Arushi Goel 07 December 2018, 09:27:46 UTC Update Readme.md 07 December 2018, 09:27:46 UTC
e3882ea Arushi Goel 07 December 2018, 09:27:29 UTC Create Readme.md 07 December 2018, 09:27:29 UTC
a7b8ac8 Xuan 07 December 2018, 09:26:50 UTC Add audio extraction code. 07 December 2018, 09:26:50 UTC
11f0974 Xuan 07 December 2018, 08:51:10 UTC Merge branch 'master' of github.com:desmond-ong/cheem-omg-empathy 07 December 2018, 08:51:10 UTC
4aac4f0 Xuan 07 December 2018, 08:50:42 UTC Rename 'combined' to 'multimodal', removed LaTeX submission files. 07 December 2018, 08:50:42 UTC
e72fe1c Arushi Goel 07 December 2018, 08:33:58 UTC Delete workspace.xml 07 December 2018, 08:33:58 UTC
a42720c Arushi Goel 07 December 2018, 08:29:24 UTC Updated test script 07 December 2018, 08:29:24 UTC
db29fdd Arushi Goel 07 December 2018, 08:21:49 UTC Adding src for visual modality 07 December 2018, 08:21:49 UTC
e1778ad Xuan 07 December 2018, 08:10:12 UTC Flatten directory structure by getting rid of src. 07 December 2018, 08:10:12 UTC
1f0ba8f Arushi Goel 07 December 2018, 08:04:31 UTC Updating train and test code 07 December 2018, 08:04:31 UTC
951fdab Arushi Goel 07 December 2018, 07:01:28 UTC Adding src for visual modality 07 December 2018, 07:01:28 UTC
2c9cb1b Xuan 07 December 2018, 06:43:24 UTC Clean-up by removing unused options. 07 December 2018, 06:43:24 UTC
4ea5009 Xuan 07 December 2018, 06:43:12 UTC Remove VRNN code to avoid clutter. 07 December 2018, 06:43:12 UTC
52365a1 Xuan 07 December 2018, 06:04:17 UTC Small fix to personalize script. 07 December 2018, 06:04:17 UTC
57e48be Xuan 07 December 2018, 03:22:00 UTC Stop hardcoding the test set into commandline options. 07 December 2018, 03:22:00 UTC
be36412 Xuan 07 December 2018, 02:33:59 UTC Add code for finetuning personalized models. 07 December 2018, 02:33:59 UTC
b975005 Xuan 07 December 2018, 02:33:42 UTC Store stories and subjects as integers, not strings. 07 December 2018, 02:33:42 UTC
65bb752 Xuan 06 December 2018, 10:02:02 UTC Import parsers into parent scripts rather than re-constructing them. 06 December 2018, 10:02:02 UTC
ff9cdbe Son Nguyen 06 December 2018, 09:48:37 UTC text-based empathy models 06 December 2018, 09:48:37 UTC
b1bf64c Xuan 06 December 2018, 03:18:06 UTC Add stories and subjects flag to cross-validation. 06 December 2018, 03:18:06 UTC
1a1120e Xuan 05 December 2018, 11:17:11 UTC Add script tocross-validate epochs. 05 December 2018, 11:17:11 UTC
93d4dce Xuan 05 December 2018, 10:59:51 UTC Save final model in train.py, add test_epoch option to crossval. 05 December 2018, 10:59:51 UTC
f6d307e Xuan 05 December 2018, 08:14:36 UTC Refactored code to extract and augment datasets. 05 December 2018, 08:14:36 UTC
83d55ba Xuan 05 December 2018, 03:58:08 UTC Fixed bug with loaded models carrying over in CV, print std when doing CV. 05 December 2018, 03:58:08 UTC
71bc092 Xuan 04 December 2018, 10:27:15 UTC Fixed some bugs, add augment_subject method to dataset code. 04 December 2018, 10:27:15 UTC
dff33b6 Xuan 04 December 2018, 09:18:26 UTC Add cross validation script for decision level fusion. 04 December 2018, 09:18:26 UTC
8b91a8d Xuan 04 December 2018, 07:12:17 UTC Made crossval save features. 04 December 2018, 07:12:17 UTC
2f81af5 Xuan 04 December 2018, 06:43:58 UTC Collate crossval predictions. 04 December 2018, 06:43:58 UTC
1179865 Xuan 03 December 2018, 16:29:29 UTC Add test_story flag to fusion script, transfer features to CPU before saving. 03 December 2018, 16:29:29 UTC
9ac375e Xuan 03 December 2018, 10:53:29 UTC Add more penalty terms to cross-validate over for SVR fusion. 03 December 2018, 10:53:29 UTC
e1ba9b9 Xuan 03 December 2018, 10:51:43 UTC Add feature saving to train.py 03 December 2018, 10:51:43 UTC
72117e1 Xuan 03 December 2018, 10:02:28 UTC Add option to output features from final FC layer. 03 December 2018, 10:02:28 UTC
66cc388 Xuan 03 December 2018, 09:56:41 UTC Enable saving predictions, compare against original annotations. 03 December 2018, 09:56:41 UTC
addeec8 Xuan 03 December 2018, 08:43:35 UTC Got SVR fusion model working. 03 December 2018, 08:43:35 UTC
632846e Xuan 03 December 2018, 07:35:35 UTC Reorganized DL-fusion code, got rid of old audio-only model files. 03 December 2018, 07:35:35 UTC
60b05d6 Xuan 03 December 2018, 07:32:01 UTC Add prediction folder option to train.py. 03 December 2018, 07:32:01 UTC
43bb815 Xuan 03 December 2018, 06:50:27 UTC Rewrote dataset code to load variable number of modalities. 03 December 2018, 06:50:27 UTC
16953e0 Xuan 02 December 2018, 15:35:00 UTC Add scaffolding code for SVR decision-level fusion. 02 December 2018, 15:35:00 UTC
ec3634b Xuan 02 December 2018, 15:04:47 UTC Add script to average predicted values. 02 December 2018, 15:04:47 UTC
46b2cbc Xuan 02 December 2018, 10:40:36 UTC Add options to normalize data and reconstruct inputs. 02 December 2018, 10:40:36 UTC
cd39a50 Xuan 02 December 2018, 09:09:34 UTC Add reconstruct flag to crossval.py 02 December 2018, 09:09:34 UTC
d7b972a Xuan 02 December 2018, 09:08:56 UTC Merge branch 'master' of github.com:desmond-ong/cheem-omg-empathy 02 December 2018, 09:08:56 UTC
0d053af Xuan 02 December 2018, 09:08:50 UTC Make sure to mask outputs, add option to reconstruct inputs. 02 December 2018, 09:08:50 UTC
b1ee59c Xuan 02 December 2018, 02:37:48 UTC More bug and formatting fixes to cross-val code. 02 December 2018, 02:37:48 UTC
4264f5e Xuan 01 December 2018, 14:52:21 UTC Fixed bugs with splitting of training data. 01 December 2018, 14:52:21 UTC
3edf052 Xuan 01 December 2018, 10:28:24 UTC Bug fixes to crossval. 01 December 2018, 10:28:24 UTC
a9ece0f Xuan 01 December 2018, 10:16:52 UTC Actually add cross-validation code. 01 December 2018, 10:16:52 UTC
5cfa7a5 Xuan 01 December 2018, 08:51:25 UTC Add training and cross validation code. 01 December 2018, 08:51:25 UTC
5311be5 Xuan 30 November 2018, 10:41:16 UTC Allow pad_shift to handle negative shifts. 30 November 2018, 10:41:16 UTC
ce82e94 Xuan 30 November 2018, 10:40:12 UTC Add code to customize train/test split. 30 November 2018, 10:40:12 UTC
80f59c8 Xuan 29 November 2018, 08:55:35 UTC Add option to specific modalities on command line for combined model. 29 November 2018, 08:55:35 UTC
3c302c2 Xuan 29 November 2018, 03:46:22 UTC Added model_dir option, change attn_len back to 3. 29 November 2018, 03:46:22 UTC
61b70c2 Xuan 28 November 2018, 08:32:27 UTC Add actor visual features as modality, make modalities optional in combined model. 28 November 2018, 08:32:27 UTC
1fe1f6d Xuan 28 November 2018, 07:28:09 UTC Make sure to sample from prior in VRNN, removed skip connections to prevent overfitting. 28 November 2018, 07:28:09 UTC
7990b54 Xuan 27 November 2018, 13:05:12 UTC Add attention on top of LSTM combined model. 27 November 2018, 13:05:12 UTC
2a7ea9f Xuan 27 November 2018, 07:14:11 UTC Add skip connections to VRNN. 27 November 2018, 07:14:11 UTC
d60468d Xuan 27 November 2018, 06:58:20 UTC Fix random seed in training code. 27 November 2018, 06:58:20 UTC
fafe5c6 Xuan 27 November 2018, 04:20:04 UTC Add annealing of loss multipliers to VRNN training. 27 November 2018, 04:20:04 UTC
ec05453 Xuan 27 November 2018, 03:19:31 UTC Move VRNN files to new folder, add eval_freq training option. 27 November 2018, 03:19:31 UTC
841b73b Xuan 26 November 2018, 11:08:49 UTC Fixed MSE calculation in VRNN training. 26 November 2018, 11:08:49 UTC
6302742 Xuan 26 November 2018, 10:36:32 UTC Got VRNN training code working. 26 November 2018, 10:36:32 UTC
717d0cf Xuan 26 November 2018, 08:09:16 UTC Add resume flag to train.py 26 November 2018, 08:09:16 UTC
e72cf2b Xuan 26 November 2018, 07:52:35 UTC Fixed how train.py calculates CCC and saves predictions. 26 November 2018, 07:52:35 UTC
780a7c7 Xuan 26 November 2018, 03:34:51 UTC Add test code to VRNN. 26 November 2018, 03:34:51 UTC
8d861d9 Xuan 25 November 2018, 11:26:45 UTC Added decoder for valence to VRNN model. 25 November 2018, 11:26:45 UTC
3219b77 Xuan 25 November 2018, 09:28:58 UTC Fixed typo with array splitting, made sure saved predictions are of same length. 25 November 2018, 09:28:58 UTC
a1c68e1 Tan Zong Xuan 25 November 2018, 08:38:44 UTC Add option to split training data into smaller sections. 25 November 2018, 08:38:44 UTC
8fc73d0 Tan Zong Xuan 25 November 2018, 04:01:50 UTC Fix ReLU typo. 25 November 2018, 04:01:50 UTC
8c57680 Tan Zong Xuan 24 November 2018, 16:33:17 UTC Add VRNN model adapted for multimodal data (no prediction yet). 24 November 2018, 16:33:17 UTC
8f47c62 Tan Zong Xuan 24 November 2018, 13:50:30 UTC Add ReLU to layers in combined model, trim saved predictions to length of original valence ratings. 24 November 2018, 13:50:30 UTC
d6f6e10 Xuan 23 November 2018, 10:41:38 UTC Fixed bug where validation was equal to training set, adding code to save predictions. 23 November 2018, 10:41:38 UTC
1dfc557 Xuan 23 November 2018, 09:47:45 UTC Added training code for combined model, fixed collate function. 23 November 2018, 09:47:45 UTC
e31ea5e Xuan 23 November 2018, 09:23:11 UTC Made collate function work for multiple modalities. 23 November 2018, 09:23:11 UTC
fdc8567 Xuan 23 November 2018, 08:41:55 UTC Add basic ATV LSTM model with feature-level fusion. 23 November 2018, 08:41:55 UTC
315c6d4 Xuan 23 November 2018, 07:56:26 UTC Add custom dataset to load combined data. 23 November 2018, 07:56:26 UTC
5d29bdf Xuan 23 November 2018, 06:39:39 UTC Reorganized paper submission template files. 23 November 2018, 06:39:39 UTC
ac5c5eb Xuan 23 November 2018, 06:35:05 UTC Merge branch 'master' of github.com:desmond-ong/cheem-omg-empathy 23 November 2018, 06:35:05 UTC
66d68a7 Xuan 23 November 2018, 06:34:52 UTC Add audio-only model. 23 November 2018, 06:34:52 UTC
back to top