Variable Bitrate Neural Fields

Takikawa, Towaki; Evans, Alex; Tremblay, Jonathan; Müller, Thomas; McGuire, Morgan; Jacobson, Alec; Fidler, Sanja

doi:10.1145/3528233.3530727

Cited by 76 publications

(16 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several methods achieve this by post-processing an existing reconstruction through incremental pruning [Deng and Tartaglione 2023] with vector quantization [Li et al 2022b]. Takikawa et al [2022] directly optimize for a compressed codebookbased representation of the scene. While these methods all report impressive compression ratios, they all rely on evaluating an MLP for each volume sample and are therefore too slow for real-time rendering of large scenes.…”

Section: Related Workmentioning

confidence: 99%

MERF: Memory-Efficient Radiance Fields for Real-time View Synthesis in Unbounded Scenes

Reiser¹,

Szeliski²,

Verbin³

et al. 2023

Preprint

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

MERF: Memory-Efficient Radiance Fields for Real-time View Synthesis in Unbounded Scenes

Reiser¹,

Szeliski²,

Verbin³

et al. 2023

Preprint

View full text Add to dashboard Cite

“…One can achieve significant training and inference speed improvements by replacing the deep multilayer perceptron with a feature voxel grid in combination with a small neural network [11,30,51] or no network at all [18,63]. Several other works achieve both fast rendering and memory-efficient storage with tensor factorizations [11], learned appearance codebooks, or quantized volumetric features [52].…”

Section: Related Workmentioning

confidence: 99%

HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling

Attal¹,

Huang²,

Richardt³

et al. 2023

Preprint

View full text Add to dashboard Cite

real-world scenes. To address these issues, we present Hy-perReel -a novel 6-DoF video representation. The two core components of HyperReel are: (1) a ray-conditioned sample prediction network that enables high-fidelity, high frame rate rendering at high resolutions and (2) a compact and memoryefficient dynamic volume representation. Our 6-DoF video pipeline achieves the best performance compared to prior and contemporary approaches in terms of visual quality with small memory requirements, while also rendering at up to 18 frames-per-second at megapixel resolution without any custom CUDA code.

show abstract

“…Since then, a flurry of works has improved various aspects of optimized neural fields, yielding higher reconstruction quality [4,75,5,70]. Neural field backbones, in particular, have become more structured and compressed [51,66,37,8,65]. The pivotal work of [45] introduced an efficient hash-based representation that allows NeRF optimizations to converge within seconds, effectively paving the way for interactive research directions on neural radiance fields.…”

Section: Related Workmentioning

confidence: 99%