@inproceedings{shrestha-etal-2026-multi,
title = "Multi-Modal-Minds@{CH}i{PSAL} 2026: A Comparative Study of Textual, Visual and Multimodal Architecture for {N}epali Meme Moderation",
author = "Shrestha, Sandesh and
K.C., Bikram and
Shah, Akshyat and
Acharya, Ashish and
Thapa, Rabin",
editor = "Sarveswaran, Kengatharaiyer and
Vaidya, Ashwini",
booktitle = "Proceedings of the Second workshop on Challenges in Processing {S}outh {A}sian Languages ({CH}i{PSAL}2026)",
month = may,
year = "2026",
address = "Palma de Mallorca, Spain",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.chipsal-1.26/",
doi = "10.63317/4ajzb6i9sfbx",
pages = "258--266",
abstract = "Memes have become ubiquitous on social media platforms blending text and imagery to express complex and culturally nuanced messages. While a high degree of automation in meme moderation has been achieved for high-resource languages, low-resource languages, such as Nepali, still remain largely neglected. In this paper, we describe our system submission to the CHiPSAL 2026 Shared Task on Multi-modal Hate and Sentiment Understanding in Low-Resource Nepali Memes, which features two main sub-tasks: (1) Detection of HateSpeech as binary classification and (2) Sentiment Analysis as multi-class classification in Nepali memes. We perform a comprehensive analysis of the following models: uni-modal textual models (mBERT, XLM-RoBERTa,MuRIL), uni-modal visual models (ResNet, ConvNeXt, ViT), nine different late-fusion multimodal models, and the vision-language foundation model, SigLIP. Among all models, the ViT model achieved the best macro F1-score(0.6278) for the hate speech detection task, while SigLIP achieved the best score (0.5481) for the sentiment analysis task. We hypothesize that the under-performance of fusion models may be attributed to OCR noise and inadequate low-resource textual representations that act as a bottleneck when paired with more advanced visual encoders. These results highlight the unique challenges of multimodal meme comprehension in low-resource contexts and underscores the requirement for culturally grounded, noise-robust approaches to content moderation in Nepali."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="shrestha-etal-2026-multi">
<titleInfo>
<title>Multi-Modal-Minds@CHiPSAL 2026: A Comparative Study of Textual, Visual and Multimodal Architecture for Nepali Meme Moderation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sandesh</namePart>
<namePart type="family">Shrestha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bikram</namePart>
<namePart type="family">K.C.</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Akshyat</namePart>
<namePart type="family">Shah</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ashish</namePart>
<namePart type="family">Acharya</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rabin</namePart>
<namePart type="family">Thapa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Second workshop on Challenges in Processing South Asian Languages (CHiPSAL2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kengatharaiyer</namePart>
<namePart type="family">Sarveswaran</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ashwini</namePart>
<namePart type="family">Vaidya</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma de Mallorca, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Memes have become ubiquitous on social media platforms blending text and imagery to express complex and culturally nuanced messages. While a high degree of automation in meme moderation has been achieved for high-resource languages, low-resource languages, such as Nepali, still remain largely neglected. In this paper, we describe our system submission to the CHiPSAL 2026 Shared Task on Multi-modal Hate and Sentiment Understanding in Low-Resource Nepali Memes, which features two main sub-tasks: (1) Detection of HateSpeech as binary classification and (2) Sentiment Analysis as multi-class classification in Nepali memes. We perform a comprehensive analysis of the following models: uni-modal textual models (mBERT, XLM-RoBERTa,MuRIL), uni-modal visual models (ResNet, ConvNeXt, ViT), nine different late-fusion multimodal models, and the vision-language foundation model, SigLIP. Among all models, the ViT model achieved the best macro F1-score(0.6278) for the hate speech detection task, while SigLIP achieved the best score (0.5481) for the sentiment analysis task. We hypothesize that the under-performance of fusion models may be attributed to OCR noise and inadequate low-resource textual representations that act as a bottleneck when paired with more advanced visual encoders. These results highlight the unique challenges of multimodal meme comprehension in low-resource contexts and underscores the requirement for culturally grounded, noise-robust approaches to content moderation in Nepali.</abstract>
<identifier type="citekey">shrestha-etal-2026-multi</identifier>
<identifier type="doi">10.63317/4ajzb6i9sfbx</identifier>
<location>
<url>https://aclanthology.org/2026.chipsal-1.26/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>258</start>
<end>266</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Multi-Modal-Minds@CHiPSAL 2026: A Comparative Study of Textual, Visual and Multimodal Architecture for Nepali Meme Moderation
%A Shrestha, Sandesh
%A K.C., Bikram
%A Shah, Akshyat
%A Acharya, Ashish
%A Thapa, Rabin
%Y Sarveswaran, Kengatharaiyer
%Y Vaidya, Ashwini
%S Proceedings of the Second workshop on Challenges in Processing South Asian Languages (CHiPSAL2026)
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma de Mallorca, Spain
%F shrestha-etal-2026-multi
%X Memes have become ubiquitous on social media platforms blending text and imagery to express complex and culturally nuanced messages. While a high degree of automation in meme moderation has been achieved for high-resource languages, low-resource languages, such as Nepali, still remain largely neglected. In this paper, we describe our system submission to the CHiPSAL 2026 Shared Task on Multi-modal Hate and Sentiment Understanding in Low-Resource Nepali Memes, which features two main sub-tasks: (1) Detection of HateSpeech as binary classification and (2) Sentiment Analysis as multi-class classification in Nepali memes. We perform a comprehensive analysis of the following models: uni-modal textual models (mBERT, XLM-RoBERTa,MuRIL), uni-modal visual models (ResNet, ConvNeXt, ViT), nine different late-fusion multimodal models, and the vision-language foundation model, SigLIP. Among all models, the ViT model achieved the best macro F1-score(0.6278) for the hate speech detection task, while SigLIP achieved the best score (0.5481) for the sentiment analysis task. We hypothesize that the under-performance of fusion models may be attributed to OCR noise and inadequate low-resource textual representations that act as a bottleneck when paired with more advanced visual encoders. These results highlight the unique challenges of multimodal meme comprehension in low-resource contexts and underscores the requirement for culturally grounded, noise-robust approaches to content moderation in Nepali.
%R 10.63317/4ajzb6i9sfbx
%U https://aclanthology.org/2026.chipsal-1.26/
%U https://doi.org/10.63317/4ajzb6i9sfbx
%P 258-266
Markdown (Informal)
[Multi-Modal-Minds@CHiPSAL 2026: A Comparative Study of Textual, Visual and Multimodal Architecture for Nepali Meme Moderation](https://aclanthology.org/2026.chipsal-1.26/) (Shrestha et al., CHiPSAL 2026)
ACL