GitHub - mihirp1998/Disentangling-3D-Prototypical-Nets: We present neural architectures that disentangle RGB-D images into objects' shapes and styles and a map of the background scene, and explore their applications for few-shot 3D object detection and few-shot concept classification.

Disentangling 3D Prototypical Networks For Few-Shot Concept Learning

Code Release- Ongoing

conda env create -f environment.yml

Download part of the dataset from this link

Update the root_location variable with the parent directory of the downloaded dataset

python main.py cs --en trainer_rgb_no_bn_munit_simple_cross_0.1_dsn

If you want me to prioritize release of something specific, then mail me at mihirp1998.mp@gmail.com

Name		Name	Last commit message	Last commit date
Latest commit History 24 Commits
archs		archs
backend		backend
experiments/default		experiments/default
images		images
lib_classes		lib_classes
nets		nets
setup		setup
DoublePool.py		DoublePool.py
README.md		README.md
cross_corr.py		cross_corr.py
environment.yml		environment.yml
exp_base.py		exp_base.py
exp_clevr_sta.py		exp_clevr_sta.py
hyperparams.py		hyperparams.py
main.py		main.py
model_base.py		model_base.py
model_clevr_sta.py		model_clevr_sta.py
pretrained_nets_carla.py		pretrained_nets_carla.py
push.sh		push.sh
split_train_test.py		split_train_test.py
tb.sh		tb.sh
tb_local.sh		tb_local.sh
utils_ap.py		utils_ap.py
utils_basic.py		utils_basic.py
utils_box.py		utils_box.py
utils_eval.py		utils_eval.py
utils_geom.py		utils_geom.py
utils_improc.py		utils_improc.py
utils_misc.py		utils_misc.py
utils_pointcloud.py		utils_pointcloud.py
utils_pyplot_vis.py		utils_pyplot_vis.py
utils_samp.py		utils_samp.py
utils_vox.py		utils_vox.py