


CMU Zhu Junyan's team developed an automatic matching ranking system to evaluate the pros and cons of various AI generation models.
This article is reproduced from Lei Feng.com. If you need to reprint, please go to the official website of Lei Feng.com to apply for authorization.
Generative AI has been very popular recently, and there are so many new pre-trained image generation models that it is dizzying to see. Whether it's portraits, landscapes, cartoons, elements of a specific artist's style, and more, each model has something it's good at generating.
With so many models, how do you quickly find the best model that can satisfy your creative desires?
Recently, Zhu Junyan, an assistant professor at Carnegie Mellon University, and others proposed the content-based model search algorithm for the first time, allowing you to search with one click Best matching depth image generation model.
##Paper address: https://arxiv.org/pdf/2210.03116.pdf
On Modelverse, an online model sharing and search platform developed by the team based on this set of model search algorithms, you can enter text, images, sketches and a given model to search for the most matching or similar related models.
Modelverse platform address: https://modelverse.cs.cmu.edu/
## Legend: input text (such as "African animals"), image (such as a landscape), sketch (such as a sketch of a standing cat) or a given model, and output the top ranking Related models (second and third rows)
For example, if you enter the text "face", you will get the following results:
##Enter a sketch of a horse:
1
Like traditional multimedia search, model search can help users find the one that best suits their specific needs model. However, the content-based model search task has its own special difficulties:
Determining whether a model can generate a specific image is a computationally difficult problem, and many deep generative models do not provide effective method to estimate density, which itself does not support the assessment of cross-modal similarity. The sampling-based method of Monte Carlo will make the model search process very slow.
To this end, Zhu Junyan’s team proposed a new model search system.
Each generative model produces an image distribution, so the authors approach the search problem as an optimization that maximizes the probability of generating a match to the query given the model. As shown in the figure below, the system consists of a pre-caching stage (a, b) and an inference stage (c).
Given a set of model, (a) first generate 50K samples for each model; (b) then encode the images into image features and calculate the first- and second-order feature statistics for each model. Statistics are cached in the system to improve efficiency; (c) during the inference phase, queries of different modalities are supported, including images, sketches, textual descriptions, another generative model, or a combination of these query types. The authors introduce an approximation here where the query is encoded as a feature vector and the model with the best similarity measure is retrieved by evaluating the similarity between the query features and each model statistics.
2 Model search effect
The author evaluates the algorithm and conducts ablation experimental analysis on 133 deep generation models (including GAN, diffusion model and autoregressive model) . Compared with the Monte Carlo baseline, this method can achieve a more efficient search, with a speed improvement of 5 times within 0.08 milliseconds, while maintaining high accuracy.
By comparing model retrieval results, we can also get a rough idea of which models can generate higher quality images for different query inputs. For example, the figure below shows the comparison of model retrieval results.
Legend: Example of model retrieval results
The top line is Image query, input still life paintings, retrieve models of related artistic styles, and get the first-ranked StyleGAN2 model and the last-ranked Vision-aided GAN model. The middle row is a sketch query, input sketches of horses and churches, and get models such as ADM and ProGAN. The bottom line is a text query. Enter "person wearing glasses" and "talking bird" to retrieve the top-ranked GANSketch model and Self-Distilled GAN model respectively.
The author also found that there are differences in model performance in different network feature spaces. As shown in the figure below, when inputting image queries, the results show that the three networks CLIP, DINO and Inception all have similar performance; when inputting sketch queries, CLIP performs significantly better, while DINO and Inception do not. Suitable for a given query, they perform better on artistic style models.
Note: Comparison of image- and sketch-based model retrieval in different network feature spaces
In addition, the model search algorithm proposed in this work can also support a variety of applications, including multi-modal user query, similar model query, real image reconstruction and editing, etc.
For example, multi-modal query can help refine the model search. When there are only images of "Nicolas Cage", only the face model can be retrieved; but when "Nicolas Cage" is also used and "dog" as input, you can retrieve the StyleGAN-NADA model that can generate the "Nicolas Cage dog" image. (As shown below)
Note: Multi-modal user query
When the input is a face model, more face generation models can be retrieved and the categories remain similar. (As shown below)
Note: Similar model query
given To determine the query image of a real face, using a higher-ranked model can obtain more accurate image reconstruction. The figure below is an example of image inverse mapping of CelebA-HQ and LSUN Church images using different ranking models.
Caption: Project the real image to the retrieved StyleGAN2 model.
#In the task of editing real images, the performance of different models also varies. In the figure below, the top-ranked image-based model retrieval algorithm is used to inversely map real images, and then edited using GANspace to turn the frowning face in the Ukiyoe image into a smiling face.
Note: Editing real images
This research has proven the feasibility of model search; There is still a lot of research space for model search for text, audio or other content generation.
But currently, the method proposed in this work still has certain limitations. For example, when querying a specific sketch, sometimes a model of an abstract shape will be matched; and sometimes when performing a multi-modal query, only a single model can be retrieved, and the system may have difficulty processing an image like a dog "elephant" ” Such multi-modal queries. (As shown below)
Note: Failure case
In addition, On its model search platform, the retrieved model list is not automatically sorted according to its effect. For example, different models are evaluated and ranked in terms of the resolution, fidelity, matching degree of the generated image, etc., so that it can be more accurate It facilitates user retrieval and can also help users better understand the pros and cons of the currently generated models. We look forward to follow-up work in this area.
The above is the detailed content of CMU Zhu Junyan's team developed an automatic matching ranking system to evaluate the pros and cons of various AI generation models.. For more information, please follow other related articles on the PHP Chinese website!

Hot AI Tools

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Undress AI Tool
Undress images for free

Clothoff.io
AI clothes remover

AI Hentai Generator
Generate AI Hentai for free.

Hot Article

Hot Tools

Notepad++7.3.1
Easy-to-use and free code editor

SublimeText3 Chinese version
Chinese version, very easy to use

Zend Studio 13.0.1
Powerful PHP integrated development environment

Dreamweaver CS6
Visual web development tools

SublimeText3 Mac version
God-level code editing software (SublimeText3)

Hot Topics

The return value types of C language function include int, float, double, char, void and pointer types. int is used to return integers, float and double are used to return floats, and char returns characters. void means that the function does not return any value. The pointer type returns the memory address, be careful to avoid memory leakage.结构体或联合体可返回多个相关数据。

The calculation of C35 is essentially combinatorial mathematics, representing the number of combinations selected from 3 of 5 elements. The calculation formula is C53 = 5! / (3! * 2!), which can be directly calculated by loops to improve efficiency and avoid overflow. In addition, understanding the nature of combinations and mastering efficient calculation methods is crucial to solving many problems in the fields of probability statistics, cryptography, algorithm design, etc.

std::unique removes adjacent duplicate elements in the container and moves them to the end, returning an iterator pointing to the first duplicate element. std::distance calculates the distance between two iterators, that is, the number of elements they point to. These two functions are useful for optimizing code and improving efficiency, but there are also some pitfalls to be paid attention to, such as: std::unique only deals with adjacent duplicate elements. std::distance is less efficient when dealing with non-random access iterators. By mastering these features and best practices, you can fully utilize the power of these two functions.

Although C and C# have similarities, they are completely different: C is a process-oriented, manual memory management, and platform-dependent language used for system programming; C# is an object-oriented, garbage collection, and platform-independent language used for desktop, web application and game development.

A function pointer is a pointer to a function, and a pointer function is a function that returns a pointer. Function pointers point to functions, used to select and execute different functions; pointer functions return pointers to variables, arrays or other functions; when using function pointers, pay attention to parameter matching and checking pointer null values; when using pointer functions, pay attention to memory management and free dynamically allocated memory; understand the differences and characteristics of the two to avoid confusion and errors.

The key elements of C function definition include: return type (defining the value returned by the function), function name (following the naming specification and determining the scope), parameter list (defining the parameter type, quantity and order accepted by the function) and function body (implementing the logic of the function). It is crucial to clarify the meaning and subtle relationship of these elements, and can help developers avoid "pits" and write more efficient and elegant code.

Algorithms are the set of instructions to solve problems, and their execution speed and memory usage vary. In programming, many algorithms are based on data search and sorting. This article will introduce several data retrieval and sorting algorithms. Linear search assumes that there is an array [20,500,10,5,100,1,50] and needs to find the number 50. The linear search algorithm checks each element in the array one by one until the target value is found or the complete array is traversed. The algorithm flowchart is as follows: The pseudo-code for linear search is as follows: Check each element: If the target value is found: Return true Return false C language implementation: #include#includeintmain(void){i

Flexible application of function pointers: use comparison functions to find the maximum value of an array. First, define the comparison function type CompareFunc, and then write the comparison function compareMax(a, b). The findMax function accepts array, array size, and comparison function parameters, and uses the comparison function to loop to compare array elements to find the maximum value. This method has strong code reusability, reflects the idea of higher-order programming, and is conducive to solving more complex problems.
