reconstruction-loss

[ROBOVIS 2026] Code for the Adaptive Keyframe Selection for Scalable 3D Scene Reconstruction in Dynamic Environments paper.

pyppur: Python Projection Pursuit Unsupervised (Dimension) Reduction To Min. Reconstruction Loss or DIstance DIstortion

The repository for the article "SSSUMO: Real-Time Semi-Supervised Submovement Decomposition". The repository is a work in progress and is going to be refactored.

In this assignment I have to build a reconstruction loss based anomaly detection model using tf.keras. I have to train an autoencoder to reconstruct Homer Simpson images, then use the trained model to classify Simpson and Flower images.

Implement a Generative Adversarial Network which is able to frontalize faces from the pose invariant features learned in our proposed pose attention guided profile to frontal network for extreme profile faces