Deployment-Aware Codec Selection for Learned RGB-D Compression in Edge--Cloud 3D Reconstruction
Abstract
Abstract Purpose: Edge--cloud multimedia pipelines couple compression, hardware execution, transport, and downstream reconstruction. In such systems, learned RGB-D codec selection cannot be reduced to an offline rate--distortion curve or to a scalar rate--distortion--complexity objective, because entropy-model seriality, runtime compatibility, and reconstruction placement are heterogeneous constraints that are not faithfully expressible as a single weighted penalty. Methods: This paper formulates codec selection as a two-stage deployment-aware decision. Aligned RGB and depth are first represented as a joint four-channel tensor, and four learned codec families are screened on SUN RGB-D in the rate--distortion--coding-time space using actual total RGB-D coded bpp and normalized-tensor PSNR. Non-dominated candidates are then filtered by explicit deployment-feasibility constraints covering bounded encoding latency, memory budget, static TensorRT runtime compatibility, entropy-model dependency structure, and edge--cloud processing placement. The selected ReLU-based factorized-prior codec is implemented on a Jetson TX2 through a hybrid TensorRT--PyTorch entropy-coding path and evaluated with RTP delivery, cloud-side decoding, and point-cloud reconstruction. Hardware-accelerated H.264/HEVC RGB-depth separate intra-coding baselines are also included under matched frame-independent coding and rate definitions. Results: The attention-enhanced model gives the strongest offline point,…
Read the paper
Related studies
- Regulation of nucleolus assembly by non-coding RNA polymerase II transcripts
- For a Healthy Histone Code, a Little SUMO in the Tail Keeps the Acetyl Away
- Discovering a binary CTCF code with a little help from BORIS
- Article Commentary: Stabilizing the Code–-Methods to Preserve RNA Prove Their Worth
- Foreword to the Special Issue for the RINTC (The Implicit Seismic Risk of Code-Conforming Structures) Project
- Commentary: Promoting Health and Well-being in Pediatric Primary Care Settings: Using Health and Behavior Codes at Routine Well-child Visits
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.