abhinavdhall Profile Banner
Abhinav Dhall Profile
Abhinav Dhall

@abhinavdhall

Followers
616
Following
3K
Media
74
Statuses
657

Academic @ Monash University

Joined May 2009
Don't wanna be here? Send us removal request.
@abhinavdhall
Abhinav Dhall
3 months
13th International Conference on Affective Computing and Intelligent Intercation (ACII) starts today at Canberra, Australia!
0
0
0
@abhinavdhall
Abhinav Dhall
4 months
Code-switching based deepfakes are non-trivial for observers. We create the first data repository to train deepfake detectors in multi-lingual settings. Collaboration with Prof. Haris Khan and Kartik @mbzuai @MonashInfotech
@mbzuai
MBZUAI
4 months
🌍 In much of the Arab world, it’s common to slip between Arabic and English in the same breath, like saying, “Yalla, let’s grab a coffee.” This linguistic blend feels natural to people, but it throws off most deepfake detectors, which are usually trained on just one language.
0
0
1
@IJCAIconf
IJCAIconf
4 months
#Workshop on Deepfake Detection, Localization and Interpretability #Keynote 🗣️Zhen Lei #CAS 🗨️Anderson Rocha @unicampoficial | Abhinav Dhall @abhinavdhall #MonashU #IJCAI2025 https://t.co/B25Eap8DT7
1
1
2
@MonashInfotech
Monash Information Technology
5 months
Great to hear A/Prof Abhinav Dhall on @SBSHindi!  Hear his expertise on whether introducing #AI in classrooms will prepare or hinder the critical thinking skills of the next generation: https://t.co/KpWGFZHKUD
0
2
5
@AcmIcmi
ACM ICMI Official
6 months
📢 Call for Workshop Papers @ ICMI 2025! Submit to one of 4 exciting workshops on: 🌏 Cross-cultural interaction (7 Jul) 🧠 Responsible affective AI (8 Jul) 🛰️ Geospatial AI + XR (15 Jul) 🤕 Pain assessment (8 Jul) Details👇 🔗 https://t.co/eL9js3XlsZ #ICMI2025 #MultimodalAI
icmi.acm.org
0
3
8
@abhinavdhall
Abhinav Dhall
7 months
Participated in a panel discussion on deepfakes hosted by @Sanjanachowhan on @NDTVProfitIndia. I stressed upon deepfakes detection and human intelligence symbosis to tackle misinformation. @MonashInfotech
@MonashInfotech
Monash Information Technology
7 months
Our expert A/Prof Abhinav Dhall was interviewed on @NDTVProfitIndia where he spoke about how India can outsmart #deepfake scams – like the one impersonating FM Nirmala Sitharaman. Watch now: https://t.co/9nyMFzK9oa @MonashUni
0
0
3
@abhinavdhall
Abhinav Dhall
7 months
We believe the resource will help learning better multi-lingual deepfake detectors. GitHub - https://t.co/7lZCJkdgHV Arxiv - https://t.co/wLHKbLUdTy Collaboration between Monash University and Mohamed bin Zayed University of Artificial Intelligence @MonashInfotech @mbzuai
0
0
2
@abhinavdhall
Abhinav Dhall
7 months
Led by @KKuckreja060702, we propose a novel LLM+GenerativeAI (lip-enactment and voice TTS/Cloning) pipeline. Authors - @KKuckreja060702, Injy Hamed, @thamar_solorio, Parul Gupta, @mharis8744 @abhinavdhall
1
1
2
@abhinavdhall
Abhinav Dhall
7 months
Yalla, let’s grab a coffee! aka يلا، تعالوا نشرب coffee! Speakers in multicultural societies commonly use language codeswitching. This becomes a big challenge from a deepfake perspective. We introduce ArEnAV dataset, the first Arabic English CSW Audio-Visual Dataset. @mharis8744
1
1
3
@MonashInfotech
Monash Information Technology
7 months
Seen a viral paraglider video circulating on your feed?  Widely published by news outlets, the video is now under scrutiny for being at least partly manipulated by #genAI. Our #Deepfakes expert @abhinavdhall weighs in for @abcnews https://t.co/F9uNgf5iSg
0
1
4
@abhinavdhall
Abhinav Dhall
7 months
Extensive evaluation of deepfakes detectors and human user study indicate the difficulty in detecting fine deepfakes in the dataset. For more details, please see https://t.co/V6mUTpTynt @parulgpt @Shreya_Ghosh_
0
2
2
@abhinavdhall
Abhinav Dhall
7 months
We evaluate the change in perception of manipulated images over multiple attributes using VLMs: viewer emotion, perceived identity, power dynamics, scene narrative, intent, and ethics possible intent of manipulation and the ethical implications of the manipulations.
1
0
0
@abhinavdhall
Abhinav Dhall
7 months
We use VLMs to reason different levels of manipulations, which effect the perception of an image in a person-centric manner. The suggestions are used to promt image manipulation VLs.
1
0
0
@abhinavdhall
Abhinav Dhall
7 months
1
0
0
@abhinavdhall
Abhinav Dhall
7 months
We introduce MultiFakeVerse image deepfakes database. Inspired by the multiverse concept: same individual can exist in multiple believable yet manipulated realities. Deepfakes aren’t just about how images are edited but what story they tell! Lead by Parul Gupta @MonashInfotech
1
1
5
@abhinavdhall
Abhinav Dhall
9 months
We are organizing the 2nd One Million Deepfake Detection Challenge at ACM Multimedia 2025, this time with an expanded dataset of over 2 million samples, making it the largest deepfake detection dataset to date. URL - https://t.co/kxYZzFZO2R Please consider participating!
0
2
6
@abhinavdhall
Abhinav Dhall
9 months
The 2nd One Million Deepfakes Detection Challenge is live as a grand challenge in ACM Multimedia 2025. Please consider participating! https://t.co/kxYZzFZO2R
0
1
2
@AcmIcmi
ACM ICMI Official
11 months
Gentle reminder that the deadline is approaching soon. Please consider submitting your workshop proposals. Workshop Chairs (icmi2025-workshop-chairs@acm.org) ========================================= Albert Ali Salah (@SzassTam ) Abhinav Dhall (@abhinavdhall )
@AcmIcmi
ACM ICMI Official
11 months
#ICMI2025 call for workshop proposal is up! Visit our website for more details: https://t.co/W46vtRaZzQ Workshop chairs: Albert Ali Salah @SzassTam Abhinav Dhall @abhinavdhall Timeline ====== Submission Deadline: February 14 2025 Notification of Acceptance: February 21 2025
0
2
4
@AcmIcmi
ACM ICMI Official
11 months
#ICMI2025 call for workshop proposal is up! Visit our website for more details: https://t.co/W46vtRaZzQ Workshop chairs: Albert Ali Salah @SzassTam Abhinav Dhall @abhinavdhall Timeline ====== Submission Deadline: February 14 2025 Notification of Acceptance: February 21 2025
icmi.acm.org
0
5
7
@TCSResearch
TCS Research
1 year
Our inaugural session of Kaleidoscope, a forum that fosters scientific collaboration & learning, featured a talk on Human-Centric Approaches for Multimodal Deepfakes Analysis by prof. @abhinavdhall, @MonashUni, @MonashInfotech at our research labs at @IITKgp. #Research #AI
2
1
8