VLDB 2026 Research / reviewers in the wild / expert
Animesh Agrawal
dblp:258/9090
· DBLP profile ↗
5ranked-venue papers
2as first author
3since 2021 · last 2025
—ORCID · conflict
Domains — the database's venue-derived domains; a paper can count in several
Artificial intelligence and machine learning · 2 · 2 since 2021Databases, data management, data science and information retrieval · 2 · 2 first-authorGraphics, computer vision, multimedia, augmented reality and games · 2 · 2 since 2021Security and privacy · 1 · 1 since 2021
Expertise — from the expertise taxonomy: the topics of the expert's papers under the CCF categories. A weight counts papers with recency: 1 for a paper about the topic, 0.3 when the topic is its context, halved every five years.
| Artificial intelligence
3 papers |
Efficient and distributed learning · 95% Trustworthy machine learning · 5% | |
| Network and information security
1 paper |
Security and privacy of machine learning · 100% |
Topics — the 8 heaviest of 10, each with the papers that count most for it
| Topic | Weight | Papers | Last | Evidence papers |
|---|---|---|---|---|
Machine learning › Efficient and distributed learning › automated machine learning › neural architecture search › one-shot neural architecture search
weight-sharing supernet |
0.9 | 1 | 2025 | VillainNet: Targeted Poisoning Attacks Against SuperNets Along the Accuracy-Latency Pareto Frontier · CCS 2025 |
Security and privacy of machine learning › adversarial attack
backdoor attack |
0.9 | 1 | 2025 | VillainNet: Targeted Poisoning Attacks Against SuperNets Along the Accuracy-Latency Pareto Frontier · CCS 2025 |
Security and privacy of machine learning › adversarial attack › backdoor attack
targeted poisoning attack |
0.9 | 1 | 2025 | VillainNet: Targeted Poisoning Attacks Against SuperNets Along the Accuracy-Latency Pareto Frontier · CCS 2025 |
Machine learning › Efficient and distributed learning
federated learning |
0.8 | 1 | 2024 | SuperFedNAS: Cost-Efficient Federated Neural Architecture Search for On-device Inference · ECCV (79) 2024 |
Machine learning › Efficient and distributed learning › federated learning › federated AutoML
federated neural architecture search |
0.8 | 1 | 2024 | SuperFedNAS: Cost-Efficient Federated Neural Architecture Search for On-device Inference · ECCV (79) 2024 |
Machine learning › Efficient and distributed learning › automated machine learning
neural architecture search |
0.8 | 1 | 2024 | SuperFedNAS: Cost-Efficient Federated Neural Architecture Search for On-device Inference · ECCV (79) 2024 |
Machine learning › Efficient and distributed learning
on-device inference |
0.8 | 1 | 2024 | SuperFedNAS: Cost-Efficient Federated Neural Architecture Search for On-device Inference · ECCV (79) 2024 |
Machine learning › Trustworthy machine learning
robustness |
0.3 | 1 | 2025 | VillainNet: Targeted Poisoning Attacks Against SuperNets Along the Accuracy-Latency Pareto Frontier · CCS 2025 |
Methods — techniques the papers use, named apart from their topics
poisoning · 1.7distance-aware optimization · 1.7neural architecture search · 0.8federated learning · 0.8
| Year | Publication | Venue | Position |
|---|---|---|---|
| 2025 | VillainNet: Targeted Poisoning Attacks Against SuperNets Along the Accuracy-Latency Pareto FrontierabstractState-of-the-art (SOTA) weight-shared SuperNets dynamically activate subnetworks at runtime, enabling robust adaptive inference under varying deployment conditions. However, we find that adversaries can take advantage of the unique training and inference paradigms of SuperNets to selectively implant backdoors that activate only within specific subnetworks, remaining dormant across billions of other subnetworks. We present VillainNet (VNET), a novel poisoning methodology that restricts backdoor activation to attacker-chosen subnetworks, tailored either to specific operational scenarios (e.g., specific vehicle speeds or weather conditions) or to specific subnetwork configurations. VNET's core innovation is a novel, distance-aware optimization process that leverages architectural and computational similarity metrics between subnetworks to ensure that backdoor activation does not occur across non-target subnetworks. This forces defenders to confront a dramatically expanded search space for backdoor detection. We show that across two SOTA SuperNets, trained on the CIFAR10 and GTSRB datasets, VNET can achieve attack success rates comparable to traditional poisoning approaches (approximately 99%), while significantly lowering the chances of attack detection, thereby stealthily hiding the attack. Consequently, defenders face increased computational burdens, requiring on average 66 (and up to 250 for highly targeted attacks) sampled subnetworks to detect the attack, implying a roughly 66-fold increase in compute cost required to test the SuperNet for backdoors. David Oygenblik, Abhinav Vemulapalli, Animesh Agrawal, Debopam Sanyal, Alexey Tumanov, Brendan Saltaformaggio |
CCS | 3 |
| 2024 | DεpS: Delayed ε-Shrinking for Faster Once-for-All Training
Aditya Annavajjala, Alind Khare, Animesh Agrawal, Igor Fedorov, Hugo Latapie, Myungjin Lee, Alexey Tumanov |
ECCV (89) | 3 |
| 2024 | SuperFedNAS: Cost-Efficient Federated Neural Architecture Search for On-device Inference
Alind Khare, Animesh Agrawal, Aditya Annavajjala, Payman Behnam, Myungjin Lee, Hugo Latapie, Alexey Tumanov |
ECCV (79) | 2 |
| 2020 | Employee Productivity and Service Quality Enhancement Using ERP With Knowledge Management: A Study in a Power Distribution Company in the Global ContextabstractMany researchers have identified that implementation of knowledge management (KM) improves the success rate of enterprise resource planning (ERP) in different sectors, but the support of KM in post-ERP implementation in the service sector is yet to be analyzed. The aim of the study is to confirm the support of KM and to identify the post-implementation effect of ERP on employees and service quality. For the fulfillment of this research, data were gathered from the service sector's employees and consumers. The research has been processed in three stages: The first stage includes the variables identification from the previously published literature and reduced to smaller groups with the help of IBM SPSS software. Second stage explores the preparation of conceptual model. The third stage involves the identification and substantiation of KM support with post-ERP implementation, which has been accomplished through a short brainstorming session with employees and senior manager of power distribution company. This research concludes that there is positive support of KM in post-ERP implementation. Animesh Agrawal, Hemant Kumar Diwakar, Suraj Kumar Mukti |
Int. J. Knowl. Manag. | 1 |
| 2020 | Knowledge Management & It's Origin, Success Factors, Planning, Tools, Applications, Barriers and Enablers: A ReviewabstractOver of the past several years, there have been rigorous discussions about the significance of knowledge management (KM) within the organization and the society. The management of knowledge is endorsed as a significant and essential factor for organizational existence and maintenance of ambitious strength. This article provides an in-depth knowledge of factors affecting KM. Literatures from 1992 to 2018 are covered in this article, 169 research papers have been explored which are related to classification of knowledge, factors affecting KM, KM tools and its planning & application. Various frameworks related to the successful implementation of KM and KM implementation tools proposed by previous authors are presented in this research article. KM is defined, classification of KM is presented, factors affecting KM are shown and its implementation strategies & tools are elucidated in available literatures in discrete manner. Animesh Agrawal, Suraj Kumar Mukti |
Int. J. Knowl. Manag. | 1 |