[CVPR2024] SchurVINS: Schur Complement-Based Lightweight Visual Inertial Navigation System

Fast Multimodal LLM on Mobile Devices

FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.

Awesome LLMs on Device: A Comprehensive Survey

CSP-J/S/X, NOIP, NOI, IOI, 信息学奥林匹克竞赛历年真题收录 | QQ交流群529507453

Unofficial implementation of PercepNet: A Perceptually-Motivated Approach for Low-Complexity, Real-Time Enhancement of Fullband Speech

[TCAD'23] AccelTran: A Sparsity-Aware Accelerator for Transformers

Manage and switch between multiple proxies quickly & easily.

🦙 LaMa Image Inpainting, Resolution-robust Large Mask Inpainting with Fourier Convolutions, WACV 2022

[ICCV 2023] MI-GAN: A Simple Baseline for Image Inpainting on Mobile Devices

🦙 LaMa Image Inpainting, Resolution-robust Large Mask Inpainting with Fourier Convolutions, WACV 2022

This is the official code for MobileSAM project that makes SAM lightweight for mobile applications and beyond!

Inpaint anything using Segment Anything and inpainting models.

BitBLAS is a library to support mixed-precision matrix multiplications, especially for quantized LLM deployment.

[ACL 2024] A novel QAT with Self-Distillation framework to enhance ultra low-bit LLMs.

Low-bit LLM inference on CPU with lookup table

Stable Diffusion and Flux in pure C/C++

[ICCV 2023] Lighting Every Darkness in Two Pairs: A Calibration-Free Pipeline for RAW Denoising && [Arxiv 2023] Make Explicit Calibration Implicit: Calibrate Denoiser Instead of the Noise Model

A list of resouces for multispectral pedestrian detection,including the datasets, methods, annotations and tools.

A Reconfigurable Accelerator with Data Reordering Support for Low-Cost On-Chip Dataflow Switching

16-bit Adder Multiplier hardware on Digilent Basys 3

Matlab implementations of algorithms and scripts of simulations presented in Informed FastICA: Semi-Blind Minimum Variance Distortionless Beamformer

The calflops is designed to calculate FLOPs、MACs and Parameters in all various neural networks, such as Linear、 CNN、 RNN、 GCN、Transformer(Bert、LlaMA etc Large Language Model)

InspireFace is a cross-platform face recognition SDK developed in C/C++, supporting multiple operating systems and various backend types for inference, such as CPU, GPU, and NPU.

The PULP Ara is a 64-bit Vector Unit, compatible with the RISC-V Vector Extension Version 1.0, working as a coprocessor to CORE-V's CVA6 core

LLM101n: Let's build a Storyteller

A novel human-interaction method for real-time speech extraction on headphones.