AIConfPaper
← Search

Anil Kag

22 accepted papers

2026

Omni-Attribute: Open-vocabulary Attribute Encoder for Visual Concept Personalization

Tsai-Shien Chen, Aliaksandr Siarohin, Gordon Guocheng Qian, Kuan-Chieh Jackson Wang, Egor Nemchinov, Moayed Haji-Ali +6
CVPR 2026

Visual concept personalization aims to transfer only specific image attributes, such as identity, expression, lighting, and style, into unseen contexts. However, existing methods rely on holistic embeddings from general-purpose image encoders, which entangle multiple visual factors and make it diffi

Cited by 0Source ↗code ↗Scholar ↗
2026

One Model, Many Budgets: Elastic Latent Interfaces for Diffusion Transformers

Moayed Haji-Ali, Willi Menapace, Ivan Skorokhodov, Dogyun Park, Anil Kag, Michael Vasilkovsky +3
CVPR 2026

Diffusion transformers (DiTs) achieve high generative quality but lock FLOPs to image resolution, limiting principled latency-quality trade-offs, and allocate computation uniformly across input spatial tokens, wasting resource allocation to unimportant regions. We introduce Elastic Latent Interface

Cited by 0Source ↗code ↗Scholar ↗
2026

SPRINT: Sparse-Dense Residual Fusion for Efficient Diffusion Transformers

Dogyun Park, Moayed Haji-Ali, Yanyu Li, Willi Menapace, Sergey Tulyakov, Hyunwoo J. Kim +2
ICLR 2026poster

Diffusion Transformers (DiTs) deliver state-of-the-art generative performance but their quadratic training cost with sequence length makes large-scale pretraining prohibitively expensive. Token dropping can reduce training cost, yet naïve strategies degrade representations, and existing methods are…

Cited by 0Source ↗code ↗Scholar ↗
2025

DenseDPO: Fine-Grained Temporal Preference Optimization for Video Diffusion Models

Ziyi Wu, Anil Kag, Ivan Skorokhodov, Willi Menapace, Ashkan Mirzaei, Igor Gilitschenski +2
NeurIPS 2025spotlight

Direct Preference Optimization (DPO) has recently been applied as a post‑training technique for text-to-video diffusion models. To obtain training data, annotators are asked to provide preferences between two videos generated from independent noise. However, this approach prohibits fine-grained comp…

Cited by 0Source ↗Scholar ↗
2025

Lightweight Predictive 3D Gaussian Splats

Junli Cao, Vidit Goel, Chaoyang Wang, Anil Kag, Ju Hu, Sergei Korolev +3
ICLR 2025poster

Recent approaches representing 3D objects and scenes using Gaussian splats show increased rendering speed across a variety of platforms and devices. While rendering such representations is indeed extremely efficient, storing and transmitting them is often prohibitively expensive. To represent large-…

Cited by 3Source ↗PDF ↗code ↗Scholar ↗
2025

Scalable Ranked Preference Optimization for Text-to-Image Generation

Shyamgopal Karthik, Huseyin Coskun, Zeynep Akata, Sergey Tulyakov, Jian Ren, Anil Kag
ICCV 2025poster

Direct Preference Optimization (DPO) has emerged as a powerful approach to align text-to-image (T2I) models with human feedback. Unfortunately, successful application of DPO to T2I models requires a huge amount of resources to collect and label large-scale datasets, e.g., millions of generated paire…

Cited by 0Source ↗PDF ↗Scholar ↗
2025

SnapGen-V: Generating a Five-Second Video within Five Seconds on a Mobile Device

Yushu Wu, Zhixing Zhang, Yanyu Li, Yanwu Xu, Anil Kag, Yang Sui +8
CVPR 2025poster

We have witnessed the unprecedented success of diffusion-based video generation over the past year. Recently proposed models from the community have wielded the power to generate cinematic and high-resolution videos with smooth motions from arbitrary input prompts. However, as a supertask of image g…

Cited by 2Source ↗PDF ↗Scholar ↗
2025

SnapGen: Taming High-Resolution Text-to-Image Models for Mobile Devices with Efficient Architectures and Training

Jierun Chen, Dongting Hu, Xijie Huang, Huseyin Coskun, Arpit Sahni, Aarush Gupta +13
CVPR 2025highlight

Existing text-to-image (T2I) diffusion models face several limitations, including large model sizes, slow runtime, and low-quality generation on mobile devices. This paper aims to address all of these challenges by developing an extremely small and fast T2I model that generates high-resolution and h…

Cited by 4Source ↗PDF ↗code ↗Scholar ↗
2025

Towards Physical Understanding in Video Generation: A 3D Point Regularization Approach

Yunuo Chen, Junli Cao, Vidit Goel, Sergei Korolev, Chenfanfu Jiang, Jian Ren +2
NeurIPS 2025poster

We present a novel video generation framework that integrates 3-dimensional geometry and dynamic awareness. To achieve this, we augment 2D videos with 3D point trajectories and align them in pixel space. The resulting 3D-aware video dataset, PointVid, is then used to fine-tune a latent diffusion mod…

Cited by 0Source ↗Scholar ↗
2024

AsCAN: Asymmetric Convolution-Attention Networks for Efficient Recognition and Generation

Anil Kag, Huseyin Coskun, Jierun Chen, Junli Cao, Willi Menapace, Aliaksandr Siarohin +2
NeurIPS 2024poster

Neural network architecture design requires making many crucial decisions. The common desiderata is that similar decisions, with little modifications, can be reused in a variety of tasks and applications. To satisfy that, architectures must provide promising latency and performance trade-offs, suppo…

Cited by 4Source ↗PDF ↗Scholar ↗
2024

BitsFusion: 1.99 bits Weight Quantization of Diffusion Model

Yang Sui, Yanyu Li, Anil Kag, Yerlan Idelbayev, Junli Cao, Ju Hu +4
NeurIPS 2024poster

Diffusion-based image generation models have achieved great success in recent years by showing the capability of synthesizing high-quality content. However, these models contain a huge number of parameters, resulting in a significantly large model size. Saving and transferring them is a major bottle…

Cited by 13Source ↗PDF ↗code ↗Scholar ↗
2024

SF-V: Single Forward Video Generation Model

Zhixing Zhang, Yanyu Li, Yushu Wu, yanwu xu, Anil Kag, Ivan Skorokhodov +6
NeurIPS 2024poster

Diffusion-based video generation models have demonstrated remarkable success in obtaining high-fidelity videos through the iterative denoising process. However, these models require multiple denoising steps during sampling, resulting in high computational costs. In this work, we propose a novel appr…

Cited by 10Source ↗PDF ↗code ↗Scholar ↗
2024

Snap Video: Scaled Spatiotemporal Transformers for Text-to-Video Synthesis

Willi Menapace, Aliaksandr Siarohin, Ivan Skorokhodov, Ekaterina Deyneka, Tsai-Shien Chen, Anil Kag +5
CVPR 2024highlight

Contemporary models for generating images show remarkable quality and versatility. Swayed by these advantages the research community repurposes them to generate videos. Since video content is highly redundant we argue that naively bringing advances of image models to the video generation domain redu…

Cited by 66Source ↗PDF ↗Scholar ↗
2024

TextCraftor: Your Text Encoder Can be Image Quality Controller

Yanyu Li, Xian Liu, Anil Kag, Ju Hu, Yerlan Idelbayev, Dhritiman Sagar +3
CVPR 2024poster

Diffusion-based text-to-image generative models e.g. Stable Diffusion have revolutionized the field of content generation enabling significant advancements in areas like image editing and video synthesis. Despite their formidable capabilities these models are not without their limitations. It is sti…

Cited by 18Source ↗PDF ↗Scholar ↗
2023

Efficient Edge Inference by Selective Query

Anil Kag, Igor Fedorov, Aditya Gangrade, Paul Whatmough, Venkatesh Saligrama
ICLR 2023poster

Edge devices provide inference on predictive tasks to many end-users. However, deploying deep neural networks that achieve state-of-the-art accuracy on these devices is infeasible due to edge resource constraints. Nevertheless, cloud-only processing, the de-facto standard, is also problematic, since…

Cited by 22Source ↗PDF ↗Scholar ↗
2021

Online Selective Classification with Limited Feedback

Aditya Gangrade, Anil Kag, Ashok Cutkosky, Venkatesh Saligrama
NeurIPS 2021spotlight

Motivated by applications to resource-limited and safety-critical domains, we study selective classification in the online learning model, wherein a predictor may abstain from classifying an instance. For example, this may model an adaptive decision to invoke more resources on this instance. Two sal…

Cited by 6Source ↗PDF ↗code ↗Scholar ↗
2021

Training Recurrent Neural Networks via Forward Propagation Through Time

Anil Kag, Venkatesh Saligrama
ICML 2021spotlight

Back-propagation through time (BPTT) has been widely used for training Recurrent Neural Networks (RNNs). BPTT updates RNN parameters on an instance by back-propagating the error in time over the entire sequence length, and as a result, leads to poor trainability due to the well-known gradient explos…

Cited by 62Source ↗PDF ↗code ↗Scholar ↗
2020

RNNs Incrementally Evolving on an Equilibrium Manifold: A Panacea for Vanishing and Exploding Gradients?

Anil Kag, Ziming Zhang, Venkatesh Saligrama
ICLR 2020poster

Recurrent neural networks (RNNs) are particularly well-suited for modeling long-term dependencies in sequential data, but are notoriously hard to train because the error backpropagated in time either vanishes or explodes at an exponential rate. While a number of works attempt to mitigate this effect…

Cited by 68Source ↗code ↗Scholar ↗