image 

image 

# Image 

 

<Gallery /> 

 

Trained on Replicate using: 

mple image.* --> 

 

### **Description** 

 

The Nemotron OCR v1 model is a state-of-the-art text recognition model designed for robust end-to-end optical character recognition (OCR) on complex real-world images. It integrates three core neural network modules: a detector for text region localization, a recognizer for transcription of detected regions, and a relational model for layout and structure analysis. 

Ovis Image 7B! 

<img src=/static-proxy?url=https%3A%2F%2Fcdn-uploads.huggingface.co%2Fproduction%2Fuploads%2F636f4c6b5d2050767e4a1491%2FcfsnngElzYv8DbTKsLohl.png width="40%"/>Enjoy! 

</div> 

mple image.* 

 

### Description 

 

The **Nemotron Page Elements v3** model is a specialized object detection model designed to identify and extract elements from document pages. While the underlying technology builds upon work from [Megvii Technology](https://github.com/Megvii-BaseDetection/YOLOX), we developed our own base model through complete retraining rather than using pre-trained weights. YOLOX is an anchor-free version of YOLO (You Only Look Once), this model combines a simpler architecture with enhanced performance. The model is trained to detect **tables**, **charts**, **infographics**, **titles**, **header/footers** and **texts** in documents. 

rong Image Tagging Model </a> and <a href="https://tag2text.github.io/">Tag2Text: Guiding Vision-Language Model via Image Tagging</a>. 

 

**Recognition and localization are two foundation computer vision tasks.** 

- **The Segment Anything Model (SAM)** excels in **localization capabilities**, while it falls short when it comes to **recognition tasks**. 

- **The Recognize Anything Model (RAM) and Tag2Text** exhibits **exceptional recognition abilities**, in terms of **both accuracy and scope**. 

### Image Upscaling Model 

 

This repository contains the PyTorch model for upscaling images. The model has been trained to upscale low-resolution images to higher resolution using convolutional neural networks. 

 

## Model Details 

Qwen Image V1 

 

This is a custom-trained **LoRA (Low-Rank Adapter)** for **Qwen Image**, fine-tuned on 85+ upscaled and varied images sourced from the infamous Hunter Biden iCloud laptop archive. Designed for **Qwen-based image generation**, this LoRA supports photorealistic and meme-style compositions for digital propaganda, viral satire, and social media chaos. Trained by [@huwhitememes](https://x.com/huwhitememes) using the [WaveSpeedAI LoRA Trainer](https://wavespeed.ai/models/wavespeed-ai/qwen-image-lora-trainer) pipeline. 

 

## 🎯 Use Cases 

side images. 

 

This is part of my submission for the Fatima Fellowship Selection Task. 

PP-HumanSeg v1 model, released by [Paddle](https://github.com/PaddlePaddle/PaddleSeg/tree/release/2.6/contrib/PP-HumanSeg). 

 

Tested on the [PP-HumanSeg-14K](https://github.com/PaddlePaddle/PaddleSeg/blob/release/2.6/contrib/PP-HumanSeg/paper.md) dataset. 

 

| Model Name | Best Input Shape | mIou(%) | Inference Time on Arm CPU(ms) | Modle Size(MB) | 

PP-HumanSeg v2 model, released by [Paddle](https://github.com/PaddlePaddle/PaddleSeg/tree/release/2.6/contrib/PP-HumanSeg). 

 

Tested on the [PP-HumanSeg-14K](https://github.com/PaddlePaddle/PaddleSeg/blob/release/2.6/contrib/PP-HumanSeg/paper.md) dataset. 

 

| Model Name | Best Input Shape | mIou(%) | Inference Time on Arm CPU(ms) | Modle Size(MB) | 

for Image Captioning using Noise-Injected CLIP](https://arxiv.org/pdf/2211.00575.pdf). 

 

Their method aims to train CLIP with only text samples. Therefore they are injecting zero-mean Gaussian Noise into the text embeddings before decoding. 

 

In their words: 

for Image Captioning using Noise-Injected CLIP](https://arxiv.org/pdf/2211.00575.pdf). 

 

Their method aims to train CLIP with only text samples. Therefore they are injecting zero-mean Gaussian Noise into the text embeddings before decoding. 

 

In their words: 

for Image Captioning using Noise-Injected CLIP](https://arxiv.org/pdf/2211.00575.pdf). 

 

Their method aims to train CLIP with only text samples. Therefore they are injecting zero-mean Gaussian Noise into the text embeddings before decoding. 

 

In their words: 

for Image Captioning using Noise-Injected CLIP](https://arxiv.org/pdf/2211.00575.pdf). 

 

Their method aims to train CLIP with only text samples. Therefore they are injecting zero-mean Gaussian Noise into the text embeddings before decoding. 

 

In their words: 

for Image Captioning using Noise-Injected CLIP](https://arxiv.org/pdf/2211.00575.pdf). 

 

Their method aims to train CLIP with only text samples. Therefore they are injecting zero-mean Gaussian Noise into the text embeddings before decoding. 

 

In their words: 

for Image Captioning using Noise-Injected CLIP](https://arxiv.org/pdf/2211.00575.pdf). 

 

Their method aims to train CLIP with only text samples. Therefore they are injecting zero-mean Gaussian Noise into the text embeddings before decoding. 

 

In their words: 

# AI-NERD 

 

AI-NERD (Artificial Intelligence for Non-Equilibrium Relaxation Dynamics) uses unsupervised deep learning to classify relaxation behavior in complex fluids directly from X-ray Photon Correlation Spectroscopy (XPCS) data. 

This repo contains a pre-trained model, which is trained on rheo-XPCS data collected at Advanced Photon Source beamline 8-ID-I. 

 

# Instance-guided Cartoon Editing with a Large-scale Dataset 

 

[](http://arxiv.org/abs/2312.01943) 

For more technical details and the codebase, please refer to our <a href="https://github.com/CartoonSegmentation/CartoonSegmentation">Github Repo</a>.