Integrating Conversational AI, Image Generation, and Code Generation:

A Uniﬁed Platform

Joylin Priya Pinto

, Mohammed Aqib, Osama Shakeel, Emad Habibi and Gagan S

Nitte Deemed to be University, NMAM Institute of Technology, Nitte, Karkala, Karnataka, India

Keywords:

DALL-E, OpenAI, API, StyleGAN, GPT-3.5-Turbo.

Abstract:

The new AI platform is designed and implemented by the software team at full-stack AI SaaS, which includes

conversational AI, image generation, and code generation. Powering for both conversation and code uses

OpenAI’s GPT-3.5- turbo and the platform will use DALL·E to generate images, thus offering a uniﬁed user

experience of these different AI villages through a single cohesive interface. Consumer-side routing is seen

in a number of technologies e.g. modern web technologies such as Next.js 13 and its App Router, thereby it

becomes possible to have smooth, fast and responsive user experience. The era in which we live is charac-

terized by user expectations that are always demanding, and this endeavor is related to vital user-experience,

developer-usability, and scalability tasks. An AI platform will only be useful if it involves the whole process

- integrating the listed elements to a solution that empowers users comprehensively across all the various do-

mains of activities will make it one. The discussion of the layout and those that use our platform is talked

about in this particular context which includes the scalability key aspect, the user interface design concern,

and the development treatment to the users. Next.js 13 together with API-s as well as the machine algorithms

of OpenAI presents the interlaid milieu of a technologic-creative realm which deals with today’s issues. Con-

sequently, we verify the efﬁciency of our system by load testing and collecting user reviews. These results not

only reafﬁrm the infallibility of our tool but also bring new areas for investigation and enhancement. In this

paper, we hope to stimulate some subtle debates on the probable of AI to make signiﬁcant changes in the user

experience and enhance the productivity of the user, moving towards a scenario in which everyone has easy

access to AI tools covering a large variety of walks of life.

1 INTRODUCTION

In the period of highly competitive races within indus-

tries competing for more efﬁciency in operations and

innovation, AI becomes a transformative technology,

not an automation, but a creative and technical phe-

nomenon in today’s industry, greatly changing work-

ﬂows and productivity in all sectors-from cus tomer

service with leading chatbots, content, and automated

code generation, transformed into simple applications

even for non-experts. With intuitive interfaces, to-

day’s users can unlock powerful AI tools without

requiring specialized programming or data science

knowledge, thus making technology more inclusive

than ever. This paper deals with the framework and

development process of an AI-powered SaaS that is

meant to bring together multiple AI functionalities to

one seamless user interface. The core AI tools in-

https://orcid.org/0000-0001-6742-7672

cluded in the platform would be conversational AI,

image generation, and code generation. We use Ope-

nAI’s GPT-3.5-turbo model for conversational AI and

code gener ation, which supports advanced natural

language processing to provide answers to questions,

give suggestions in decision- making scenarios, and

generate code upon prompts by the user. The fea-

ture of image generation is supported by OpenAI’s

DALL·E API on the platform, whereby unique im-

ages can be produced based on text descriptions;

therefore, it opens greater possibilities for artists, de-

signers, and general users on this creative front. The

platform architecture is based upon Next.js 13, which

is a new web framework with support for server-side

rendering and client-side routing, providing ﬁne scal-

ability and responsive performance. This is the ﬁrst

platform with ﬂuent navigation due to the use of App

Router in Next.js 13, and simpler modular, scalable

development due to the nature of React with compo-

174

Pinto, J. P., Aqib, M., Shakeel, O., Habibi, E. and S, G.

Integrating Conversational AI, Image Generation, and Code Generation: A Uniﬁed Platform.

DOI: 10.5220/0013588800004664

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 3rd International Conference on Futuristic Technology (INCOFT 2025) - Volume 2, pages 174-180

ISBN: 978-989-758-763-4

nents. Tailwind CSS enhances the looks of the plat-

form with responsive designs to create visually ap-

pealing and user- friendly interfaces across many de-

vices. We included Clerk to manage user authentica-

tion, thereby adding security and ensuring a seamless

engagement with users. This AI SaaS platform devel-

opment underlines the need for performance, acces-

sibility, and innovation in obtaining for de velopers,

creators, and even casual users, advanced AI models.

We show that sophisticated AI can civilize both tech-

nical and creative challenges in an accessible, user-

centered manner by combining OpenAI’s GPT-3.5-

turbo and DALL·E APIs with an efﬁcient Next.js 13

framework. It sets a benchmark for bringing AI closer

to the mainstream and transforming the user experi-

ence in powerful and practical ways.

2 LITERATURE REVIEW

Rapid advances in artiﬁcial intelligence (AI) have

transformed industries by making it possible to

perform activities like natural language processing

(NLP), image generation, and code generation that

were previously thought to be the sole purview of hu-

man intelligence. A major factor in this change has

been the introduction of AI models such as GPT- 3

and DALL·E, which have features that mimic human-

like text comprehension and production as well as

incredibly life- like text-to-image synthesis. AI-

powered SaaS platforms are revolutionizing enter-

prises by increasing efﬁciency, fostering innovation,

and optimizing processes. This analysis highlights re-

search gaps and their implications for the future of

AI solutions by examining signiﬁcant advancements,

integration difﬁculties, and user- focused design in

these platforms.

Harshil T. Kanakia et.al. (Kanakia and Nair, 2023)

suggested an approach that com-bines server-side and

client-side technology to create an application that

generates images using OpenAI’s Image GPT. The

React-built user interface lets users enter text to cre-

ate images, which are then handled by a server that

runs on Node.js and Express.js and processes the data

via an API. In order to create pixel values via a de-

coder utilizing autoregressive approaches, the server

communicates with OpenAI’s Image GPT, which to-

kenizes (using Byte Pair Encoding), embeds, and pro-

cesses through a Transformer network. A smooth

and effective system for text-to-image synthesis is en-

sured by storing the produced images, which are se-

mantically aligned with the input text, in a MongoDB

database for later retrieval. Following picture gener-

ation, the client receives an API response from the

server that contains the created image, which is sub-

sequently saved in a MongoDB database via Mon-

goose.js.

Amon Rapp et.al (Rapp et al., 2025) examined how

individuals react to images produced by a Gen-AI

text-to-image model, pay- ing particular attention

to the impressions, evaluations, and feelings they

arouse. It also aimed to investigate whether people

are capable of coming up with logical ”folk theories”

regarding how these technologies function internally.

Twenty participants in the study, who were given par-

ticular written prompts and twenty distinct visuals

produced by Stable Dif- fusion, participated in semi-

structured interviews. Heat maps were used to exam-

ine participant experiences and determine which input

features had the most impact on the visuals. The study

aimed to investigate how the general population views

AI-generated images outside of direct interaction with

the technology and emphasized the signiﬁcance of

visual signals in comprehending Gen-AI’s behavior.

They looked at how individuals responded to Gen-AI-

generated photos, speciﬁcally how they evaluated the

images’ representation and aesthetic quality. The ma-

jority of participants thought the Gen- AI photos were

”strange” or ”prototypical,” which made them unset-

tling and made them consider the biases in the image

creation process. Participants reworked their percep-

tions and understandings by either overvaluing or un-

dervaluing the Gen- AI in an attempt to lessen their

experience of unfamiliarity. Many respondents cited

the themes of strangeness and self- overvaluation as

the most prevalent, even though some saw the images

as having utilitarian or epistemological worth.

Aditya Ramesh et al. (Ramesh et al., 2022) out-

lines a generative model that combines a prior and a

decoder to create visuals from text. The prior uses

diffusion and autoregressive (AR) techniques to cre-

ate CLIP image embedding’s zi from text captions y.

These embedding’s are converted into images by the

decoder using a modiﬁed diffusion model. In order

to enhance image quality during training, the authors

also cover the use of classiﬁer- free guidance. Two

up samplers are utilized to produce high- resolution

images, and PCA is used to decrease the CLIP image

embedding’s in order to preserve important informa-

tion during training. The diffusion prior uses a Trans-

former model to directly predict the un-noised pic-

ture embedding, whereas the AR prior is conditioned

on both the text caption and the CLIP text embed-

ding. Contrastive models like CLIP have been shown

to learn robust representations of images that capture

both semantics and style. Moreover, the joint embed-

ding space of CLIP enables language-guided image

Integrating Conversational AI, Image Generation, and Code Generation: A Uniﬁed Platform

175

manipulations in a zero- shot fashion. We use diffu-

sion models for the decoder and experiment with both

auto-regressive and diffusion models for the prior,

ﬁnding that the latter are computationally more efﬁ-

cient and produce higher-quality samples.

Paula Maddigan et. al. (Maddigan and Susnjak, 2023)

highlighted that the implementation of practical NLIs

has always been difﬁcult because natural language is

inherently ambiguous and because badly written user

queries make it difﬁcult for language models to de-

termine user intent. This paper suggests using the

devel- opments in pre-trained large language mod-

els (LLMs), like ChatGPT and GPT-3, to transform

free-form natural language straight into code for suit-

able visualisations, as an alternative to the conven-

tional approach of creating new language model itera-

tions. This study introduces Chat2VIS, a unique sys-

tem that leverages LLM capabilities and shows how

efﬁcient rapid engineering can handle the challenging

challenge of language comprehension more quickly,

leading to simpler and more accurate in contrast to

the previous study.

Arghavan Moradi Dakhel et. al. (Dakhel et al., 2023)

presented an example of how the OpenAI Codex

might be used as an AI code assistant. It assists de-

velopers by suggesting lines or blocks of code in real

time, which makes the developers more productive

since they don’t have to write repetitive code. Nev-

ertheless, Copilot sometimes doesn’t have contex-

tual understanding especially when coding is complex

and needs a lot of clariﬁcation for recommendations.

Genius addresses all those shortcomings by combin-

ing conversational AI with code generation that has

been made more contextual and intuitive for coding as

needs and preferences of the developer are concerned.

Tero Karras at. el. (Karras, 2019) provided a dif-

ferent generator architecture for adversarial genera-

tive networks. The style-based architecture allows

for ﬁne control over the images generated. Another

advancement is StyleGAN2, which further enhanced

image quality and resulted in very realistic outputs

with fewer artifacts. These developments have estab-

lished new standards for photorealism in the gener-

ation of AI-driven images. But, however, the mod-

els are extremely rigid with minimal in- teraction and

customizations. The authors Genius capitalize from

here, to add personalization in their version of DALL-

E’s image generation capabilities which allows users

to create images customized for any particular need-

an advantage that leaves every other technology to the

bench.

3 SYSTEM DESIGN AND

ARCHITECTURE

Our platform brings together three leading-edge AI

capability conversational AI, image generation, and

code generation within a single easy-to-use interface.

While each individually has strong performance, their

coming together in one uniﬁed platform provides a

one-of-a-kind seamless experience. Such an inte-

grated design enables diverse users - developers and

content creators, to entrepreneurs and casual users-

to use AI in wide-ranging applications without difﬁ-

culty. The result is intuitive, accessible, and powerful;

even the most complex AI conversations ﬂow seam-

lessly. Figure 1 shows the architecture of the proposed

method.

Figure 1: Architecture of the Proposed Method

3.1 Conversational AI with

GPT-3.5-turbo

We have deployed OpenAI’s GPT-3.5-turbo model,

one of the leading cutting-edge language models

which guarantees human-like interactions. Unlike the

regular chatbot that uses already written response,

GPT-3.5-turbo relies on deep learn- ing to decipher

tricky prompts and respond with contextual richness.

This allows for a much more ﬂexible and engaging in-

teraction where the AI can keep pace with varied user

inquiries, engage in back-and-forth exchanges, and

adjust responses based on the ﬂow of conversation.

For users, this conversational AI tool would behave

INCOFT 2025 - International Conference on Futuristic Technology

176

like a personal assistant, capable of handling techni-

cal questions, brainstorming new ideas, and providing

quick research insights or even managing customer

service inquiries. Businesses that wish to inject some

life into AI-powered assistants in customer support

could use this model towards creating a more natu-

ral and non-robotic feel for the conversational expe-

rience. Beyond that, the AI could prove useful for

developers and technical users by being a support-

ive tool for any project-related insights or code assis-

tance, making the conversational AI component one

of the most versatile tools available on the platform.

3.2 Use of DALL·E in Image Generation

Our platform features the AI model DALL·E, which

was trained to produce high-quality original images

from text- based prompts. DALL·E’s power is that it

can really decode complex and imaginative text de-

scriptions, allowing users to see what they want with-

out having design skills. Imagine describing the “sun-

set over a futuristic city skyline with neon-lit build-

ings,” and DALL·E brings it to life in seconds. This

has huge potential for digital marketers, content cre-

ators, visual artists, and anyone trying to produce cap-

tivating visuals speedily. Our platform makes possi-

ble access to such an advanced image-generating tool,

allowing for creative discovery as well as the ability

to quickly prototype ideas or create content for social

media, marketing campaigns, and so on. This capa-

bility further democratizes the creative process, where

people with ideas but perhaps less professional-level

design experience can create high-value visual con-

tent. DALL·E’s direct integration brings any abstract

ideas ﬁrmly into tangible, sharable formats- this both

ampliﬁes productivity and creativity.

3.3 GPT-3.5-turbo for Code Generation

The ﬁnal piece of our stack is code generation, driven

by GPT-3.5-turbo. Using this capability the end-

user types a natural-language command, and then the

model translates that into executable code in many

programming languages. This functionality goes a

long way toward really empowering developers to

quickly prototype and automate vast quantities of of-

ten tedious, hand-coding work. The ability to un-

derstand natural language also makes it accessible to

users who are not technically qualiﬁed, and one can

now create functional code without much program-

ming experience.

For the developers, it acts as an intelligent assistant

in coding; it absorbs repetitive tasks, generates boil-

erplate code, or even suggests more complex logic.

Suppose a user types in “write a Python function to

calculate the factorial of a number” and, in a matter

of seconds, has readable, efﬁcient code. This is not

completing code; this tool will then respond to clari-

fying questions and provide options while it evolves

changes to the output in response to the user’s feed-

back. It relates the natural language to code and helps

transform users’ ideas into working code with a lot

of speed to achieve development and improve project

outcomes.

4 IMPLEMENTATION

Implementation of our platform will involve dif-

ferent layers ranging from the front end to the back

end through API integration. Such elements are inte-

gral for us to get the users using the technology with

as smooth a joyride as possible.

4.1 Front-End Development

The front-end was developed with a huge emphasis

put on user experience and functionality. With the

use of React, a component- based architecture allows

for the reusability of UI elements across the applica-

tion. Therefore, common components such as but-

tons, forms, and modals would be utilized efﬁciently

throughout sections of the platform, ensuring design

and functionality consistency. All of these AI mod-

els, in short, a conversational model, an image gener-

ation model, a code-generation model have their own

pages, all orchestrated through the powerful capabil-

ities of Next.js 13’s App Router. This has enhanced

routing. Not only this but has also reduced loading

times, so users can get just the functionalities they

need to use without waiting in vain. In order to easily

access our site, we made use of Tailwind CSS for the

responsive design. Using this ensures that the look

and functionality of our site look beautiful on any de-

vice. Our users can ensure a smooth experience using

their desktop computer at home or in a cafe´ while

running their ofﬁce.

4.1.1 Conversational Model

In the proposed architecture, GPT-3.5-turbo model is

integrated to provide human-like conversations across

various use cases:

• Chatbot interactions: Assists users with real-time

responses and natural language processing.

• Customer support: Acts as a virtual assistant to an-

swer queries.

Integrating Conversational AI, Image Generation, and Code Generation: A Uniﬁed Platform

177

• Creative brainstorming: Helps users generate ideas,

summaries, and insights.

• Backend: The conversational API is handled

through a Next.js server that processes user inputs and

fetches responses from OpenAI.

Figure 2 shows the working of the conversational

model of the proposed system.

Figure 2: Working of Conversational Model

4.1.2 Image Generation Model

• The platform integrates DALL·E API to generate

creative images from user-provided text prompts.

• Use cases: Designing custom artwork, generating

content visuals, or prototyping. Here, user inputs a

text prompt through the front-end UI. Backend routes

the request to the DALL·E API.

• The generated image is displayed to the user via the

front end.

• Performance: The system ensures quick responses

with proper error handling for invalid inputs.

This feature provides users with creative image gen-

eration capabilities, making it useful for content cre-

ators and designers. Figure 3 shows the working of

the image generation model of the proposed system.

Figure 3: Working of Image Generation Model

4.1.3 Code Generation Model

• The GPT-3.5-turbo model powers the code genera-

tion feature to assist developers with code snippets.

Generating ready-to-use code for speciﬁc tasks (e.g.,

sorting algorithms, API calls). Bug ﬁxes and sugges-

tions help users debug code or optimize logic.

• Language support: Covers multiple programming

languages such as Python, Java, and JavaScript.

• Front-end input ﬁelds where users describe the re-

quired code. API requests sent to the GPT-3.5-turbo

backend to generate the corresponding code snippet.

• The result is displayed with syntax highlighting for

better readability. This feature improves productiv-

ity by giving developers quick access to relevant code

solutions. Fig. 4 shows the working of the conversa-

tional model of the proposed system.

Figure 4: Working of Conversational Model

4.2 Back-End Development

We’ve used Node.js and Express for the back-end

part of our application. Pretty much everything that

was required in order to handle API requests on

the OpenAI advanced models being GPT-3.5-turbo

and DALL·E was given. Another point made on

the performance-oriented back-end structure was the

data fetching provided by Next.js 13 for real-time re-

sponses in front- end components.

It goes a long way in ensuring that as soon as users

send in their requests-whether it is generating text,

images, or code- the responses are served fast and cor-

rect. Our back-end architecture is optimized to take

simultaneous requests just in case users will be mak-

ing loads of requests at one time, thereby making the

interaction more responsive.

4.3 Performance Review

To measure our system, we carried out thorough per-

formance testing, which involved test cases on load

strength, quality of the response served by the AI,

and user experience generally. User experience is

exactly what we measure at the heart. From experi-

ments and through feedback from users, we gathered

information on how people interact with our web-

site. Users were clearly satisﬁed with our website.

However, they were most satisﬁed with the conversa-

tional AI feature. Users opined that the answers re-

ceived from the system were always very speciﬁc and

relevant. Interacting with the system remained both

funny and informative. What was most pleasantly

surprising was that in the case of image generation

through DALL·E, the feedback was highly positive.

Users were astonished at the capacity of the model

to generate visual images on text prompts that var-

ied incredibly and even exceeded their expectations.

This ability not only enhances the creative process but

also opens new avenues for the generation of digital

content. The code generation feature also left devel-

opers a game- changer, as users pointed out how it

streamlined their workﬂow in terms of avoiding a lot

of redundant coding. Saves time, as well as increases

productivity, letting developers look at more complex

and innovative aspects of a project rather than getting

bogged down with nitty-gritty coding issues. Overall,

the performance review conﬁrms our platform’s suc-

cessful fulﬁllment of all necessary user needs since it

is a combination of state-of- the-art technology and a

INCOFT 2025 - International Conference on Futuristic Technology

178

user-friendly design to constitute a powerful tool for

quite a diversiﬁed range of applications.

5 CHALLENGES FACED AND

SOLUTIONS

5.1 Integration of AI Tools

Challenge: Integrating the GPT-3.5-turbo and

DALL·E APIs into a single platform presented com-

patibility issues. Each API had its own authentication

methods, rate limits, and data handling requirements,

complicating the seamless interaction needed for real-

time user experiences.

Solution: To address this challenge, a uniﬁed back-

end service was developed using Node.js and Express,

which streamlined API interactions. A middleware

layer was created to handle authentication and re-

quests to the OpenAI APIs. This abstraction layer

simpliﬁed the process of communicating with both

tools, allowing for efﬁcient data management and er-

ror handling. Furthermore, extensive testing was per-

formed to optimize API call efﬁciency and minimize

latency.

5.2 User Authentication and Security

Challenge: Secure user authentication was a key re-

quirement to safeguard user data and ensure a safe

user experience. The system design was complex as

it had to manage user sessions, authentication tokens,

and sensitive data without compromising security.

Solution: Clerk’s authentication solution was inte-

grated to the platform, providing a solid framework

for user management. Secure sign-up and login were

provided, along with token-based authentication for

session management. Best practices for securing API

endpoints and encrypting sensitive information were

followed to ensure that user data were secure.

5.2.1 Performance Optimization

Challenge: This made performance and response

times signiﬁcant issues since the multiple functional-

ities on the platform also comprised various AI tools.

Sloppy loading and sluggish response could be disas-

trous for users, especially on live applications.

Solution: Different approaches are also used for per-

formance optimization. The server-side rendering in

Next.js reduces the initial load times while using the

application front end. Other than that, the response

from the APIs is also cached which prevents their call.

Loads have been tested so that gives more room for

reﬁnement on both the front and the back-end.

5.3 User Interface Design

Challenge: The key was to make it easy to understand

and intuitive and communicate the potential of the

platform. Balancing functionality with aesthetics in

order to satisfy many varied user needs required very

careful design considerations. Solution: With this

progression, user feedback was asked to be included

throughout the process; hence, usability testing intro-

duced iterative reﬁnements in that design. For lay-

outs, Tailwind CSS has been adopted, making the de-

signs re- sponsive as well as attractive by applying a

modular, easier interface using React components to

develop it. Always having access standards in mind

so as to make it possible to the wider Audience’s us-

ability also applied.

6 CONCLUSION AND FUTURE

WORK

Conclusively, our Software as a Service (SaaS) plat-

form has reached all the targets set out in its roadmap

version. It is a very successful platform that inte-

grates the GPT-3.5- turbo technology of OpenAI and

the DALLE API, which is the most recent invention of

the GPT family. The platform, moreover, not only en-

sures real-life conversations mimicking the users but

also through this facilitates the process of image gen-

eration through textual descriptions and code creation

from speciﬁc prompts. From the beginning, our main

goal was to create a comprehensive solution for de-

velopers, content creators, and businesses. As a re-

sult of our careful planning and execution, we have

turned our vision into a reality. The user-centric expe-

rience that was offered through the use of Next.js 13

for server-side rendering and the implementations of

React and Tailwind CSS in the front end has enabled

us to accomplish this task. Additional security mea-

sures such as the use of Clerk for user authentication

have increased the reliability of the platform, thereby

protecting data while creating a feeling of trust among

users.

In addition, the modular architecture that made the

platform easy to extend and adapt to user needs

and technological advancement were the other advan-

tages. Adding extra features with AI capabilities such

as sentiment analysis and advanced NLP tools can

greatly be of beneﬁt to the users. Sentiment analy-

sis would enable us to look at the interactions on a

much more granular level, while advanced NLP tools

Integrating Conversational AI, Image Generation, and Code Generation: A Uniﬁed Platform

179

could also allow for context awareness and nuance in

conversation to further enhance user engagement and

satisfaction. The main expansion of use cases is sig-

niﬁcant to the growth of the platform, as the poten-

tial areas of application involve personalized educa-

tion tools that use conversational AI for tutoring, cre-

ative design applications for collaborative image gen-

eration, or automated coding assistants for develop-

ers. That means in order to increase market relevance

and adoption, we tailor the platform to speciﬁc indus-

tries, including e-commerce, healthcare, and enter-

tainment. An effective feedback mechanism will help

induce the users to share their experience and provide

suggestions directly through the platform. In such a

user- driven approach, it might be possible to guide

updates and development of future features so that the

platform develops accordingly considering the needs

and expectations of users. We believe that dedication

to continuous improvement and a focus on user expe-

rience will make a powerful tool to empower users in

creative or professional endeavors.

REFERENCES

Dakhel, A. M., Majdinasab, V., Nikanjam, A., Khomh, F.,

Desmarais, M. C., and Jiang, Z. M. J. (2023). Github

copilot ai pair programmer: Asset or liability? Jour-

nal of Systems and Software, 203:111734.

Kanakia, H. T. and Nair, S. P. (2023). Designing a user-

friendly and responsive ai based image generation

website and performing diversity assessment of the

generated images. In 2023 4th International Confer-

ence on Electronics and Sustainable Communication

Systems (ICESC), pages 1090–1097. IEEE.

Karras, T. (2019). A style-based generator architecture

for generative adversarial networks. arXiv preprint

arXiv:1812.04948.

Maddigan, P. and Susnjak, T. (2023). Chat2vis: generating

data visualizations via natural language using chatgpt,

codex and gpt-3 large language models. Ieee Access,

11:45181–45193.

Ramesh, A., Dhariwal, P., Nichol, A., Chu, C., and

Chen, M. (2022). Hierarchical text-conditional im-

age generation with clip latents. arXiv preprint

arXiv:2204.06125, 1(2):3.

Rapp, A., Di Lodovico, C., Torrielli, F., and Di Caro, L.

(2025). How do people experience the images cre-

ated by generative artiﬁcial intelligence? an explo-

ration of people’s perceptions, appraisals, and emo-

tions related to a gen-ai text-to-image model and its

creations. International Journal of Human-Computer

Studies, 193:103375.

INCOFT 2025 - International Conference on Futuristic Technology

180