Search Results for "fusion"
Sort By:
Showing 31 open source projects for "fusion"
View related business solutions-
Custom VMs From 1 to 96 vCPUs With 99.95% UptimeLive migration and automatic failover keep workloads online through maintenance. One free e2-micro VM every month.
-
Build Agents and Models on One PlatformGemini Enterprise Agent Platform is Google Cloud's comprehensive platform for developers to build, scale, govern, and optimize agents and models. Choose from Google's most advanced models and third-party models like Anthropic's Claude Model Family.
-
1
UForm
Multi-Modal Neural Networks for Semantic Search, based on Mid-Fusion
...Early-fusion models encode both modalities jointly so they can take into account fine-grained features. Usually, these models are used for re-ranking relatively small retrieval results. Mid-fusion models are the golden midpoint between the previous two types. Mid-fusion models consist of two parts – unimodal and multimodal. -
2
ViMax
Director, Screenwriter, Producer, and Video Generator All-in-One
...The system aims to bridge foundational vision backbones and generative language models through adapters and fusion layers that maximize both signal integration and reasoning depth, and includes utility pipelines for training, evaluation, and deployment. -
3
Multimodal
TorchMultimodal is a PyTorch library
This project, also known as TorchMultimodal, is a PyTorch library for building, training, and experimenting with multimodal, multi-task models at scale. The library provides modular building blocks such as encoders, fusion modules, loss functions, and transformations that support combining modalities (vision, text, audio, etc.) in unified architectures. It includes a collection of ready model classes—like ALBEF, CLIP, BLIP-2, COCA, FLAVA, MDETR, and Omnivore—that serve as reference implementations you can adopt or adapt. The design emphasizes composability: you can mix and match encoder, fusion, and decoder components rather than starting from monolithic models. ... -
4
HunyuanCustom
Multimodal-Driven Architecture for Customized Video Generation
...It supports conditioning via image, audio, video, and text, and can perform subject replacement in videos, generate avatars speaking given audio, or combine multiple subject images. The architecture builds on HunyuanVideo, with added modules for identity reinforcement and modality-specific condition injection. Text-image fusion module based on LLaVA for improved multimodal understanding. Applicable to single- and multi-subject scenarios, video editing/replacement, singing avatars etc. -
Train ML Models With SQL You Already KnowBuild and deploy ML models using familiar SQL. Automate data prep with built-in Gemini. Query 1 TB and store 10 GB free monthly.
-
5
MESHROOM
3D reconstruction software
...The dense modeling of the scene is the result yielded by chaining two computer vision-based pipelines, “Structure-from-Motion” (SfM) and “Multi View Stereo” (MVS). Fusion of Multi-bracketing LDR images into HDR. Alignment of panorama images. Support for fisheye optics. Automatically estimate fisheye circle or manually edit it. Take advantage of motorized-head file. Easy to integrate in your Renderfarm System. Add specific rules to select the most suitable machines regarding CPU, RAM, GPU requirements of each Node. -
6
Advanced RAG Techniques
Advanced techniques for RAG systems
...It includes hands-on Jupyter notebooks and runnable scripts that show how to implement ideas like optimizing chunk sizes, proposition chunking, HyDE/HyPE query transformations, fusion retrieval, reranking, and ensemble retrieval. There is also an evaluation section that demonstrates how to measure RAG performance and compare different configurations in a systematic way. -
7
DLRM
An implementation of a deep learning recommendation model (DLRM)
...It includes data loaders for standard benchmarks (like Criteo), training scripts, evaluation tools, and capabilities like mixed precision, gradient compression, and memory fusion to maximize throughput. -
8
VMZ (Video Model Zoo)
VMZ: Model Zoo for Video Modeling
The codebase was designed to help researchers and practitioners quickly reproduce FAIR’s results and leverage robust pre-trained backbones for downstream tasks. It also integrates Gradient Blending, an audio-visual modeling method that fuses modalities effectively (available in the Caffe2 implementation). Although VMZ is now archived and no longer actively maintained, it remains a valuable reference for understanding early large-scale video model training, transfer learning, and multimodal... -
9
Grounded-Segment-Anything
Marrying Grounding DINO with Segment Anything & Stable Diffusion
...The core idea behind the project is to pair Grounding DINO — a zero-shot object detector that can locate objects described by natural language — with Segment Anything Model (SAM), which can produce detailed masks for objects once they are localized. This fusion lets users provide arbitrary text descriptions (e.g., “a cat, a bicycle, or a coffee mug”), have the detection model find relevant bounding boxes, and then use SAM to generate precise segmentation masks that isolate each object in the scene. -
$300 Free Credits to Build on Google CloudPut your $300 in credit toward real workloads, then keep building with free monthly usage for 20+ products. No commitment and no charge until you upgrade.
-
10
Claw Compactor
14-stage Fusion Pipeline for LLM token compression
Claw Compactor is a utility designed to optimize and manage the context limitations inherent in AI agent systems, particularly those built on OpenClaw-like architectures. It addresses the challenge of finite context windows in language models by compressing or summarizing historical interactions while preserving essential information. The system works by transforming older conversation data into condensed representations that maintain continuity without exceeding token limits. This approach... -
11
Omnilingual ASR
Omnilingual ASR Open-Source Multilingual SpeechRecognition
Omnilingual-ASR is a research codebase exploring automatic speech recognition that generalizes across a very large number of languages using shared modeling and training recipes. It focuses on leveraging self-supervised audio pretraining and scalable fine-tuning so low-resource languages can benefit from high-resource data. The project provides data preparation pipelines, training scripts, decoding utilities, and evaluation tools so researchers can reproduce results and extend to new... -
12
PoJamas aims to provide a Python and tools for loading, processing, and producing .cr2, pz3 (crz, pzz) files compatible with the SmithMicro (e-frontier) Poser character animation application. PoJamas is composed of: - Python library - Python Wavefront (.obj) 3D viewer based on GLFW - LibreOffice/Python Application (to ease the library and the viewer usage) As of 2020, the project is ported in Python3 As of 2021 this project proposes a 3D viewer for Wavefront files...
-
13
Make-A-Video - Pytorch (wip)
Implementation of Make-A-Video, new SOTA text to video generator
Implementation of Make-A-Video, new SOTA text to video generator from Meta AI, in Pytorch. They combine pseudo-3d convolutions (axial convolutions) and temporal attention and show much better temporal fusion. The pseudo-3d convolutions isn't a new concept. It has been explored before in other contexts, say for protein contact prediction as "dimensional hybrid residual networks". The gist of the paper comes down to, take a SOTA text-to-image model (here they use DALL-E2, but the same learning points would easily apply to Imagen), make a few minor modifications for attention across time and other ways to skimp on the compute cost, do frame interpolation correctly, get a great video model out. ... -
14
Patch-NetVLAD
Multi-Scale Fusion of Locally-Global Descriptors for Place Recognition
This repository contains code for the CVPR2021 paper "Patch-NetVLAD: Multi-Scale Fusion of Locally-Global Descriptors for Place Recognition". -
15
ViReMa
Viral Recombination Mapper
ViReMa (Viral Recombination Mapper) detects and reports recombination or fusion events in virus genomes using deep sequencing datasets. Feb 2014 - Our paper (Open Access) is available at Nucleic Acids Research: "Discovery of functional genomic motifs in viruses with ViReMa–a Virus Recombination Mapper–for analysis of next-generation sequencing data" http://nar.oxfordjournals.org/content/42/2/e11 This is an on-going project and updates will be regularly posted. -
16
NanoDet-Plus
Lightweight anchor-free object detection model
...In NanoDet-Plus, we propose a novel label assignment strategy with a simple assign guidance module (AGM) and a dynamic soft label assigner (DSLA) to solve the optimal label assignment problem in lightweight model training. We also introduce a light feature pyramid called Ghost-PAN to enhance multi-layer feature fusion. These improvements boost previous NanoDet's detection accuracy by 7 mAP on COCO dataset. NanoDet provide multi-backend C++ demo including ncnn, OpenVINO and MNN. There is also an Android demo based on ncnn library. Supports various backends including ncnn, MNN and OpenVINO. Also provide Android demo based on ncnn inference framework. -
17
Mask2Former
Code release for "Masked-attention Mask Transformer
Mask2Former is a unified segmentation architecture that handles semantic, instance, and panoptic segmentation with one model and one training recipe. Its core idea is to cast segmentation as mask classification: a transformer decoder predicts a set of mask queries, each with an associated class score, eliminating the need for task-specific heads. A pixel decoder fuses multi-scale features and feeds masked attention in the transformer so each query focuses computation on its current spatial... -
18
U-Net Fusion RFI
U-Net for RFI Detection based on @jakeret's implementation
...This project will use the aoflagger program within the code, so you may need to ensure that any environment variables are set for aoflagger before use. cite: https://sourceforge.net/p/u-net-fusion-rfi/wiki/cite/ -
19
Earth Enterprise
Google Earth Enterprise - Open Source
...Unlike Google Maps or Google Earth, Earth Enterprise does not include Google’s proprietary imagery but instead provides the tools needed to manage and visualize private geospatial datasets. The system is composed of three main components: Fusion, which processes and integrates imagery, vector, and terrain data into unified map layers; Server, which hosts the resulting globes or maps via Apache or Tornado-based web servers; and Client, which includes the Google Earth Enterprise Client (EC) for 3D visualization and the Google Maps JavaScript API V3 for 2D viewing. Designed for enterprise, research, and government use, it allows for secure, scalable deployment of geospatial visualization systems within private infrastructure. -
20
FusionCatcher
Somatic fusion-genes finder for RNA-seq data
FusionCatcher searches for novel/known somatic fusion genes, translocations, and chimeras in RNA-seq data (paired-end reads from Illumina NGS platforms like Solexa and HiSeq) from diseased samples. The aims of FusionCatcher are: - very good detection rate for finding candidate fusion genes, - very easy to use (i.e. no a priori knowledge of databases and bioinformatics is needed in order to run FusionCatcher), - very good detection of challenging fusion genes, like for example IGH fusions, CIC fusions, DUX4 fusions, CRLF2 fusions, TCF3 fusions, etc... -
21
BasicSR
Winning Solution in NTIRE19 Challenges on Video Restoration
...Unlike single-image restoration models, EDVR addresses the temporal dimension by aligning multiple video frames using deformable convolutional layers in a coarse-to-fine manner, allowing it to effectively handle large motion and complex scene dynamics. The architecture includes bespoke modules (e.g., Pyramid, Cascading and Deformable alignment and Temporal Spatial Attention fusion) that align information across frames and emphasize important features for restoration, enabling state-of-the-art performance on benchmarks such as the REDS challenge. By fusing spatial and temporal context, EDVR produces significantly improved visual quality in restored videos compared with approaches that treat each frame independently. -
22
Makani
Makani was developed a commercial-scale airborne wind turbine
...The software enables simulation, control, and analysis of the Makani M600 turbine system, designed to operate offshore and autonomously manage complex aerodynamic behaviors. Built primarily in C++ and Python, the codebase integrates real-time flight control, sensor fusion, aerodynamic modeling, and telemetry visualization. The project also provides comprehensive simulation environments for studying airborne wind power systems and flight dynamics. Although Makani’s commercial development concluded, the software remains valuable for researchers and engineers. -
23
RefineNet
RefineNet: Multi-Path Refinement Networks
RefineNet is a MATLAB-based framework for semantic image segmentation and general dense prediction tasks. It implements the architecture presented in the CVPR 2017 paper RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation and its extended version published in TPAMI 2019. The framework uses multi-path refinement and improved residual pooling to achieve high-quality segmentation results across multiple benchmark datasets. It provides trained models for datasets... -
24
IPS Framework
A simple Python framework for loosely-coupled multiphysics simulations
The IPS Framework has migrated to https://github.com/HPC-SimTools/IPS-framework The IPS framework was developed to support loosely-coupled integrated modeling of fusion plasmas (and was originally the Integrated Plasma Simulator). However, the design is quite general, and is suitable for many scientific and engineering domains. In addition to plasma physics, it is also being used in the engineering of batteries. One of the novel features of the IPS framework is its ability to support parallelism at multiple levels: components can launch individual parallel tasks, and also launch multiple tasks concurrently. ... -
25
htmlarea
Small, powerful, full featured WYSIWYG editor
HTMLArea 4 is a browser based WYSIWYG editor that easily replaces the TEXTAREA in your web pages. It is written in JavaScript, and suitable for use in any modern web browser, and any page on your web site. Current version is 4.0-2016-08-29
