mlsanityData to 14 Oct 2026

Parallel Multiscale Autoregressive Density Estimation

arXiv 1703.03664 ↗First posted 10 Mar 2017cs.CV cs.NEcited by 71

Not on the map. Why?

Abstract (arXiv)

PixelCNN achieves state-of-the-art results in density estimation for natural images. Although training is fast, inference is costly, requiring one network evaluation per pixel; O(N) for N pixels. This can be sped up by caching activations, but still involves generating each pixel sequentially. In this work, we propose a parallelized PixelCNN that allows more efficient inference by modeling certain pixel groups as conditionally independent. Our new PixelCNN model achieves competitive density estimation and orders of magnitude speedup - O(log N) sampling instead of O(N) - enabling the practical generation of 512x512 images. We evaluate the model on class-conditional image generation, text-to-image synthesis, and action-conditional video generation, showing that our model achieves the best results among non-pixel-autoregressive density models that allow efficient sampling.

Key facts. Cited by 71 papers in the map (data through 14 Oct 2026). First posted on arXiv 10 Mar 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 25; 2022: 14; 2023: 8; 2024: 5; 2025: 13; 2026: 6.

71 papers.

Sort:

By area:

  1. 2024Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Predictioncited by 1,023
  2. 2021Generating Images with Sparse Representationscited by 266
  3. 2021Cross-Modal Contrastive Learning for Text-to-Image Generationcited by 259
  4. 2021Deep Generative Modelling: A Comparative Review of VAEs, GANs, Normalizing Flows, Energy-Based and Autoregressive Modelscited by 252
  5. 2021Autoregressive Diffusion Modelscited by 210
  6. 2022Generative Modelling With Inverse Heat Dissipationcited by 167
  7. 2023Neural Video Compression with Diverse Contextscited by 165
  8. 2022Reinforcement Learning with Action-Free Pre-Training from Videoscited by 108
  9. 2021FitVid: Overfitting in Pixel-Level Video Predictioncited by 87
  10. 2025Unified Multimodal Understanding and Generation Models: Advances, Challenges, and Opportunitiescited by 85
  11. 2021PixelSynth: Generating a 3D-Consistent Experience from a Single Imagecited by 82
  12. 2023ARTboldsymbol·V: Auto-Regressive Text-to-Video Generation with Diffusion Modelscited by 76
38 more
  1. 20213D Semantic Scene Completion: a Surveycited by 74
  2. 2021Learning to Compose Visual Relationscited by 72
  3. 2021Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codescited by 71
  4. 2021Combiner: Full Attention Transformer with Sparse Computation Costcited by 64
  5. 2021Predicting Video with VQVAEcited by 61
  6. 2022Look Outside the Room: Synthesizing A Consistent Long-Term 3D Scene Video from A Single Imagecited by 55
  7. 2021Adversarial Text-to-Image Synthesis: A Reviewcited by 52
  8. 2024Zero-Shot Detection of AI-Generated Imagescited by 50
  9. 2021DAE-GAN: Dynamic Aspect-aware GAN for Text-to-Image Synthesiscited by 45
  10. 2024Autoregressive Models in Vision: A Surveycited by 39
  11. 2023A Survey on Graph Diffusion Models: Generative AI in Science for Molecule, Protein and Materialcited by 35
  12. 2022Image Compression with Product Quantized Masked Image Modelingcited by 33
  13. 2022Deep Lossy Plus Residual Coding for Lossless and Near-lossless Image Compressioncited by 22
  14. 2022HARP: Autoregressive Latent Video Prediction with High-Fidelity Image Generatorcited by 22
  15. 2023Vector Quantized Wasserstein Auto-Encodercited by 22
  16. 2021Multiscale deep context modeling for lossless point cloud geometry compressioncited by 21
  17. 2025RestoreVAR: Visual Autoregressive Generation for All-in-One Image Restorationcited by 15
  18. 2025Fine-Tuning Visual Autoregressive Models for Subject-Driven Generationcited by 13
  19. 2021LC-FDNet: Learned Lossless Image Compression with Frequency Decomposition Networkcited by 12
  20. 2022Diffuser: Efficient Transformers with Multi-hop Attention Diffusion for Long Sequencescited by 12
  21. 2025Uniform Discrete Diffusion with Metric Path for Video Generationcited by 12
  22. 2021M3L: Language-based Video Editing via Multi-Modal Multi-Level Transformerscited by 8
  23. 2022ECM-OPCC: Efficient Context Model for Octree-based Point Cloud Compressioncited by 8
  24. 2026Evolution of Video Generative Foundationscited by 7
  25. 2023Restoration based Generative Modelscited by 6
  26. 2023HiGen: Hierarchical Graph Generative Networkscited by 6
  27. 2022Split Hierarchical Variational Compressioncited by 5
  28. 2022Vision-Language Matching for Text-to-Image Synthesis via Generative Adversarial Networkscited by 5
  29. 2021Multi-Resolution Continuous Normalizing Flowscited by 4
  30. 2022Parallel Neural Local Lossless Compressioncited by 4
  31. 2022Practical Learned Lossless JPEG Recompression with Multi-Level Cross-Channel Entropy Model in the DCT Domaincited by 4
  32. 2022Long-horizon video prediction using a dynamic latent hierarchycited by 4
  33. 2021Hierarchical Video Generation for Complex Datacited by 3
  34. 2021Conditional COT-GAN for Video Prediction with Kernel Smoothingcited by 3
  35. 2021Pretrained equivariant features improve unsupervised landmark discoverycited by 2
  36. 2021CAGAN: Text-To-Image Generation with Combined Attention GANscited by 2
  37. 2021Inverting and Understanding Object Detectorscited by 2
  38. 2024An Efficient Implicit Neural Representation Image Codec Based on Mixed Autoregressive Model for Low-Complexity Decodingcited by 2

50 most cited of 71 All 71 (JSON)