This is the artifact for paper “Automated Testing of Software that Uses Machine Learning APIs (#158)” in ICSE2022

[ICCV 2021 (Oral Presentation)] Dual-Camera Super-Resolution with Aligned Attention Modules (RefSR)

CCNet: Criss-Cross Attention for Semantic Segmentation (TPAMI 2020 & ICCV 2019).

Datasets, Transforms and Models specific to Computer Vision

[CVPR 2022] Official CoTTA Code for our paper Continual Test-Time Domain Adaptation

Adaptive Video Streaming with Layered Neural Codecs

A real-time streaming framework that utilizes neural network-based super-resolution to enhance live video quality

🔥🔥🔥 Latest works on video streaming/processing/analysis

[BMVC'23] Prompting Visual-Language Models for Dynamic Facial Expression Recognition

Survey Paper List - Efficient LLM and Foundation Models

FPGA-based hardware acceleration for dropout-based Bayesian Neural Networks.

Miro[ACM MobiCom '23] Cost-effective On-device Continual Learning over Memory Hierarchy with Miro

Simulation of Rhythmic pixel regions. Encodes and decodes frames using region information.

A curated list of early exiting (LLM, CV, NLP, etc)

This is originally a collection of papers on neural network accelerators. Now it's more like my selection of research on deep learning and computer architecture.

prema simulation

An Android application to serve as a prototype for MobiSpectral

Source code and datasets for Ekya, a system for continuous learning on the edge.

[CVPR 2019, Oral] HAQ: Hardware-Aware Automated Quantization with Mixed Precision

[ICLR 2020] Once for All: Train One Network and Specialize it for Efficient Deployment

[NeurIPS 2022] Official code for "Focal Modulation Networks"

A co-design architecture on sparse attention

PyTorch implementation of "Supervised Contrastive Learning" (and SimCLR incidentally)