Optimized and restructured BASNet model and loss #67

DheerajMadda · 2023-09-07T04:36:42Z

Optimized the loss function, removed For-Loop for calculating IoU and used torchmetrics for SSIM.
Restructured the BASNet model definition for better readability + removed unnecessary computations of side outputs (Sup1, Sup2, Sup3, Sup4, Sup5, Sup6) for model in "Eval Mode".
Created a BASNet_Lite model with MobileNetv3-Large encoder, that follows the restructured script same as 2).

BASNet Base (Resnet34 encoder) Information:
Model Size = ~332 MB
Number of Parameters = ~87 Million
FLOPs = ~195 GFLOPs

BASNet Lite (MobileNetv3-Large encoder) Information:
Model Size = ~15 MB
Number of Parameters = ~3.9 Million
FLOPs = ~4.1 GFLOPs

1) Optimized the loss function, removed for loop for calculating IoU and used torchmetrics for SSIM. 2) Restructured the BASNet model definition 3) Created a BASNet_Lite model

DheerajMadda added 4 commits September 7, 2023 09:53

Optimized and restructured BASNet model and loss

a285562

1) Optimized the loss function, removed for loop for calculating IoU and used torchmetrics for SSIM. 2) Restructured the BASNet model definition 3) Created a BASNet_Lite model

Update __init__.py

666f71c

Update BASNet.py

21bc4bd

Update BASNet_lite.py

9f49efe

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Optimized and restructured BASNet model and loss #67

Optimized and restructured BASNet model and loss #67

DheerajMadda commented Sep 7, 2023 •

edited

Loading

Optimized and restructured BASNet model and loss #67

Are you sure you want to change the base?

Optimized and restructured BASNet model and loss #67

Conversation

DheerajMadda commented Sep 7, 2023 • edited Loading

DheerajMadda commented Sep 7, 2023 •

edited

Loading