Awesome

VCONV_DAE_TF

VCONV_DAE for Tensorflow

• Reference Paper : https://arxiv.org/abs/1604.03755

• Reference GitHub Repository(Torch & Matlab) : https://github.com/Not-IITian/VCONV-DAE

• Current Training Result of Reconstruction on ModelNet40 (30x30x30), Val_Loss(1.14E-05) <br><img src="Img/bathtub_te_1_input.png" width="450px" height="300px"></img><br/> <br><img src="Img/bathtub_te_1_output.png" width="450px" height="300px"></img><br/> <br><img src="Img/chair_te_1_input.png" width="450px" height="300px"></img><br/> <br><img src="Img/chair_te_1_output.png" width="450px" height="300px"></img><br/>

• Current Training Result of Reconstruction on KNU_Simplification (Downsampled to 30x30x30), Val_Loss (0.18) <br><img src="Img/KNU_Downsampled/COMP13920_Before.png" width="450px" height="300px"></img><br/> <br><img src="Img/KNU_Downsampled/COMP13920_Before_Predict.png" width="450px" height="300px"></img><br/> <br><img src="Img/KNU_Downsampled/Backdoor4_After.png" width="450px" height="300px"></img><br/> <br><img src="Img/KNU_Downsampled/Backdoor4_After_Predict.png" width="450px" height="300px"></img><br/>