Coding & Development

SubSeq is a software company dedicated to building and scaling software products through a repeatable product engine. The company's core mission is to develop innovative software solutions and then scale those that resonate most with users. While the specific AI functionalities mentioned in the stale description (automating tasks like writing tickets, planning sprints, and updating timelines across platforms like Jira, Slack, and GitHub) are not explicitly detailed on the current live website, the company's focus on a 'repeatable product engine' suggests a systematic and potentially automated approach to product development and growth. The website emphasizes building and scaling products that users love, indicating a user-centric and market-driven strategy.

Deep-voice-conversion is an open-source project implemented in TensorFlow, designed for voice style transfer using deep neural networks. This tool enables users to convert a source voice to a specific target voice, notably demonstrated with the voice of actress Kate Winslet. A key differentiator is its ability to perform voice conversion without requiring parallel data (like source and target voice recordings of the same utterance), relying instead on a collection of target speaker waveforms and a small set of <wav, phone> pairs from anonymous speakers. The architecture comprises two main modules: Net1 for phoneme classification and Net2 for speech synthesis, utilizing CBHG modules for feature extraction from sequential data. It's ideal for researchers and developers interested in advanced voice manipulation techniques.

motion-diffusion-model is an open-source PyTorch implementation of the "Human Motion Diffusion Model" paper, designed for generating and editing human motion sequences. The tool boasts significant speed improvements, now running 40X faster with a 50-diffusion-step model and optimized CLIP calling. It supports various tasks including text-to-motion, action-to-motion, and unconstrained motion synthesis. Users can generate motions from text prompts or actions, render SMPL meshes, and perform motion editing such as in-between and upper-body modifications. The project also integrates DiP for ultra-fast text-to-motion and offers features like DistilBERT text encoder support and dataset caching for faster loading.

Dia Industries is at the forefront of robotics and AI, focusing on transforming work through innovation with their humanoid robots. Their flagship product, Alve Generation 01, is a physical humanoid robot designed for controlled tasks at scale, integrating advanced AI with precise mechanical engineering. The company offers a comprehensive developer platform for dataset management, AI model development, and deployment in industrial settings. Additionally, their Alve AI Technology provides an advanced AI platform for industrial process optimization, predictive maintenance, and automated decision-making across manufacturing environments. Key applications include manufacturing for precision assembly and quality control, logistics for warehouse automation, and quality assurance through automated inspection and compliance verification.

DeepLabCut is an open-source toolbox designed for state-of-the-art markerless pose estimation across various animals and humans. It leverages deep learning to track user-defined features, making it highly versatile and applicable to a wide range of behaviors and species. The tool provides a user-friendly GUI and API, integrating advanced models and frameworks while offering sensible defaults for life scientists. It supports both single and multi-animal pose estimation, identification, and tracking. DeepLabCut is actively maintained, offering continuous improvements, including faster performance variants, real-time capabilities, and a recent backend migration to PyTorch for enhanced flexibility and easier installation. Comprehensive documentation, online courses, and a model zoo are available to assist users.

Gapier provides a platform with free Actions APIs specifically designed for GPT creators. This tool empowers developers to significantly expand the capabilities of ChatGPT by facilitating the integration of various external services. It is particularly beneficial for those looking to create custom GPT actions and enhance the overall functionality and interactivity of their AI applications. Gapier aims to bridge the gap between AI models and real-world services, making AI more versatile.

Netify offers comprehensive network intelligence through high-performance Deep Packet Inspection (DPI). The platform provides full transparency into network activities by identifying applications, extracting traffic metadata, and leveraging real-time data feeds. It combines local deep packet inspection with network intelligence to deliver business-driven results. Netify's solutions are designed for cybersecurity and analytics, enabling users to understand what's happening on their network. Key offerings include the Netify DPI Agent for local inspection, Netify Data Feeds for real-time information, and Netify Informatics for advanced analytics, supporting various use cases from application identification to Tor IP data analysis.

Deep Learning Streamer (DL Streamer) Pipeline Framework is an open-source streaming media analytics framework built on the GStreamer multimedia framework. It enables the creation of complex media analytics pipelines for both cloud and edge deployments. DL Streamer is optimized for performance and functional interoperability across GStreamer plugins, supporting various backend libraries. It utilizes OpenVINO™ for inference on Intel CPU, GPU, and VPU platforms, VA-API for GPU-accelerated video decode/encode, and OpenCV/DPC++ for image processing. The framework supports a wide range of models including VLMs, object detection, classification, and human pose detection, making it suitable for diverse applications like retail analytics, industrial inspection, and security monitoring.

DiscoFaceGAN is a TensorFlow-based implementation for disentangled and controllable face image generation, as presented in a CVPR 2020 Oral paper. This tool allows for the creation of virtual people's faces with precise control over identity, expression, pose, and illumination. It achieves this through 3D imitative-contrastive learning, embedding 3D priors into adversarial learning to imitate the image formation of a 3D face deformation and rendering process. A key feature is its factor disentanglement, ensuring that changing one factor (e.g., expression) does not affect others. The tool also supports reference-based generation, real image pose manipulation, lighting editing, and expression transfer, making it valuable for researchers and developers working with facial image synthesis and manipulation.

eRPC (Embedded RPC) is an open-source Remote Procedure Call (RPC) system specifically designed for multichip embedded systems and heterogeneous multicore SoCs. Unlike other modern RPC systems, eRPC distinguishes itself by being optimized for tightly coupled systems, utilizing plain C for remote functions, and maintaining a small code footprint (less than 5kB). It is not intended for high-performance distributed systems over a network. eRPC allows developers to export existing C functions without significant prototype changes, and includes a code generator tool, erpcgen, which accepts IDL files to generate shim code for serialization and invocation in C/C++ or Python. It supports various transports like CMSIS UART, NXP Kinetis SPI, TCP/IP, and USB CDC, making it versatile for different embedded environments.

OBBDetection is an open-source oriented object detection library built upon MMdetection v2.2, designed for researchers and developers working with object detection tasks. It inherits all features from MMdetection, ensuring a robust and familiar environment. The library supports multiple frameworks and implements various oriented object detectors like RoI Transformer and Gliding Vertex. A key feature is its flexible representation of oriented boxes, accommodating Horizontal Bounding Boxes (HBB), Oriented Bounding Boxes (OBB), and 4-point boxes (POLY). It leverages BboxToolkit for oriented bounding box operations and includes a model zoo with benchmarks for supported methods and backbones. The project is released under the Apache 2.0 license.

ESPCN offers a PyTorch implementation of the Efficient Sub-Pixel Convolutional Neural Network, designed for real-time single image and video super-resolution. Based on a CVPR 2016 paper, this tool allows users to upscale images and videos with various factors (2x, 3x, 4x, 8x). It includes scripts for training and testing, with support for datasets like VOC2012 for training and various benchmark datasets for testing. The implementation provides benchmarks for different upscale factors and demonstrates image and video results, making it valuable for researchers and developers in image processing and computer vision.

LucidInside.com is an exclusive domain available for sale on the Odys Marketplace, targeting businesses in VR/AR technology and digital marketing. The offering includes the domain name, complimentary logo design, and built-in SEO benefits from an existing authority backlink profile. The domain, aged 6 years (since 2018) with a DR of 34 and over 117 unique referring domains, aims to save buyers significant time and money on SEO and branding. The purchase process is streamlined into three steps: choosing the domain, transferring ownership, and downloading brand assets. Testimonials highlight Odys Global's reliability and the quality of their aged domains for SEO and marketing strategies.

EasyClangComplete is a powerful and easy-to-use plugin designed for Sublime Text 3 and 4, offering comprehensive code completion capabilities for C, C++, Objective-C, and Objective-C++. It aims to enhance coding efficiency and reduce errors for developers working with these languages. The plugin supports various configuration methods, including automatic integration with CMake projects and Bazel compilation databases (on Linux and macOS). For those not using CMake or Bazel, extensive documentation is provided to guide users through alternative configuration options. As an open-source project, EasyClangComplete encourages community contributions, allowing users to report issues, suggest features, and fix bugs.

usulnet is a self-hosted Docker management platform designed for DevOps and SysAdmins, offering a robust UI to manage Docker infrastructure. It provides full lifecycle control over containers, stacks, and Docker Swarm clusters, along with multi-node management via a master/agent architecture. Key features include Trivy CVE security scanning, real-time monitoring with alerts, a built-in Nginx reverse proxy with automatic HTTPS, and an embedded authoritative DNS server. The platform also offers a web-based crontab manager, browser-based RDP/VNC access, and database/LDAP browsers. With authentication features like TOTP 2FA and RBAC, usulnet aims to be a comprehensive, open-source alternative to tools like Portainer, deployable in 60 seconds.

PET (Pattern-Exploiting Training) is an open-source research tool designed for few-shot text classification and natural language inference. It employs a semi-supervised training procedure that reformulates input examples as cloze-style phrases, allowing language models to better understand tasks. The tool, along with its iterative variant iPET, demonstrates significant performance improvements over traditional supervised training and other semi-supervised baselines, even surpassing GPT-3 in some low-resource scenarios while requiring substantially fewer parameters. It supports various training modes including PET, iPET, and supervised training, and offers evaluation methods like unsupervised and priming. Researchers can use PET for 13 different tasks, including SuperGLUE tasks, and can also customize it for their own specific applications by defining DataProcessors and PVPs (Pattern-Verbalizer Pairs).

Fast-SRGAN is an open-source deep learning model designed for real-time super-resolution, enabling the upsampling of low-resolution videos to high resolution at 30 frames per second. Built on the SR-GAN architecture and utilizing pixel shuffle for speed, this tool is ideal for enhancing video quality efficiently. It includes a pre-trained generator model on the DIV2k dataset, featuring 8 residual blocks and 64 filters. Users can easily run inference on their own images or train the model with custom settings via a configurable YAML file and command-line parameters. The project provides speed benchmarks, demonstrating its capability to upsample to 720p at around 30fps on an M1 Pro GPU. It also offers clear instructions for installation, usage, and training, making it accessible for developers and researchers.

MakeTheDocs streamlines the documentation process by leveraging AI to create comprehensive documentation pages from demonstration videos. Users simply upload a video, and the AI analyzes it to generate a full documentation page. This tool is designed to save time and effort, allowing teams to focus on innovation rather than manual documentation. Key features include fast AI analysis and generation, the ability to add company branding and set documentation goals, and storage for previous generations. MakeTheDocs supports exporting documentation in various formats like TXT, PDF, and Markdown, and offers different plans based on video length and token usage.

AIgentor is a public AI character chat platform designed for fast, no-login access to AI conversations. Users can browse a wide array of characters, start chatting instantly, and even create their own public characters anonymously without needing a traditional account. The platform prioritizes a low-friction experience, offering cleaner navigation and faster loading times compared to other character chat sites. It is built around free access, supported by advertising, and aims to make character chat feel immediate by minimizing unnecessary steps or forced account systems. Character creation is reviewed and moderated to ensure content suitability and safety.

Facial-Expression-Recognition is an open-source deep learning project built with TensorFlow, designed for real-time facial detection in video streams and subsequent recognition of emotional expressions. The tool leverages trained models that have achieved 65% accuracy on the fer2013 dataset, making it a valuable resource for researchers and developers in the field of computer vision and emotion AI. It is primarily tested on Ubuntu and macOS Sierra, offering a robust solution for these environments. Users can easily run a demo to capture faces via webcam and recognize expressions, or train their own models from scratch by downloading and integrating the fer2013 dataset. The project is dependent on Python (>= 3.3), TensorFlow (>= 1.1.0), and OpenCV, providing a clear pathway for installation and usage.

Face-Pose-Net provides a DCNN model and Python code for robustly estimating 6 degrees of freedom (6DoF) 3D face pose or 11 parameters of a 3x4 projection matrix from unconstrained images. A key differentiator is its ability to perform face alignment without relying on fragile landmark detectors, making it highly effective even with low-resolution, occluded, or near-profile views. The tool integrates with a Face Renderer to create an end-to-end pipeline for facial pose estimation and generating multiple rendered views for alignment and data augmentation. It supports both CPU and GPU for extremely fast pose estimation and offers improved face recognition through better face alignment compared to state-of-the-art landmark detectors.

few-shot-object-detection (FsDet) offers official implementations of few-shot object detection benchmarks, including the ICML 2020 paper "Frustratingly Simple Few-Shot Object Detection." It introduces new benchmarks across PASCAL VOC, COCO, and LVIS datasets, with multiple groups of few-shot training examples and evaluation results for both base and novel classes. The repository provides benchmark results and pre-trained models for a two-stage fine-tuning approach (TFA), where the detector is first trained on abundant base classes and then fine-tuned on a small balanced training set. FsDet is modular, allowing for easy integration of custom datasets and models, serving as a general framework for future research in few-shot object detection.

PassGAN is an open-source deep learning tool for password guessing, implementing the approach described in the paper "PassGAN: A Deep Learning Approach for Password Guessing." This repository provides a modified TensorFlow implementation of Improved Training of Wasserstein GANs, making it easy to train and sample from the model. It includes a command-line interface for generating password samples and training custom models. A pretrained PassGAN model, trained on the RockYou dataset, is also provided. Users can train their own models using various password leaks and datasets, with instructions for downloading common datasets like the LinkedIn leak. The tool is released under an MIT License, acknowledging the original authors of the PassGAN paper and the underlying WGAN training code.

pnnx (PyTorch Neural Network eXchange) is an open standard designed for PyTorch model interoperability. It offers an open model format for PyTorch, meticulously defining computation graphs and high-level operators to strictly match PyTorch's architecture. The tool enables users to optimize their PyTorch models, reduce dependencies on extension packages, and convert models between various formats like TorchScript, ONNX, and NCNN. pnnx also facilitates the export of models to a portable pnnx format, ONNX-zero, or NCNN, making them suitable for deployment on different platforms. It supports both Python pip installation and portable binary packages, offering flexibility for developers.