C3: High-Fidelity Neural Video/Image Compression

DeepMind's per-instance overfit neural codec achieving low decode complexity at comparable rate-distortion to standard baselines, with published code.

Link

• Visit resource

Related resources

• SVT-JPEG-XSIntel's SVT JPEG XS codec implementation (ISO/IEC 21122), low-latency encoder/decoder for broadcast and AR use cases.
• A Complete End-to-End Open Source Toolchain for VVCAcademic paper describing complete open-source VVC (H.266) encode/decode toolchain including VVenC/VVdeC.
• xin26xFrom-scratch HEVC/H.26x encoder written in plain C, claims 4x speed vs x265 single-core.
• jxs (JPEG XS reference)Reference software implementation of ISO/IEC 21122-5 (JPEG XS) including jxs_encoder binary, supports PNM/PGX/raw YUV/RGB input.
• SVT-AV1 documentation treeCanonical SVT-AV1 docs covering encoder parameters, build guide, encoder design and per-tool appendices for CDEF, CfL, film grain, global m…
• Xilinx/AMD Video SDK — on-prem getting started (Alveo U30)Official docs for deploying the Alveo U30 video acceleration stack: XRT, the XRM resource manager and FFmpeg/GStreamer plugins with the XMA…
• nyanmisaka/ffmpeg-rockchipFFmpeg fork supporting Rockchip rkmpp hardware decode/encode and RGA scaling, Jellyfin-focused for RK35xx SoCs.
• avm (AOM Video Model) — AV2 reference codecAOMedia's reference encoder/decoder for the next-gen AV2 codec (successor to AV1); experimental research codebase with first 1.0.0 tag.
• Qencode Transcoding API referenceREST API documentation for Qencode cloud transcoding covering HLS/DASH/MP4 outputs, H.264/H.265/VP9, DRM, thumbnails, subtitles and custom…
• Making the rav1d Video Decoder 1% FasterDetailed performance teardown of the Rust rav1d AV1 decoder versus C dav1d, walking through profiling and codegen-level fixes.
• SVT-AV1 Deep Dive — preset/CRF benchmark methodologyLarge-scale SVT-AV1 benchmark (thousands of encodes) across presets and CRF using SSIMULACRA2 and XPSNR with documented harmonic-mean metho…