580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How do advanced lossless image compression algorithms balance compression ratio, encoding/decoding speed, and applicability to different image types?
This research area focuses on evaluating and comparing state-of-the-art lossless image compression algorithms to understand trade-offs among compression efficiency, computational complexity (encoding/decoding times), and performance across various image types (e.g., 8-bit/16-bit grayscale and RGB). Such insights are critical in selecting appropriate algorithms for diverse applications ranging from medical imaging to multimedia transmission where both speed and fidelity are paramount.
Key finding: This paper provides a comprehensive evaluation of top lossless image compression standards—such as lossless JPEG, PNG, JPEG 2000, JPEG-LS, WebP, FLIF, and AVIF—across 8-bit and 16-bit grayscale and RGB images. It quantifies... Read more
Key finding: This work systematically analyzes foundational lossless image compression techniques including run-length coding, entropy coding (Shannon-Fano, Huffman, arithmetic), and dictionary-based methods (LZW). It clarifies the... Read more
Key finding: This study proposes a hybrid lossless compression framework combining LZW algorithm with repeated application of Bose, Chaudhuri and Hocquenghem (BCH) error-correcting codes to boost compression ratio without data loss.... Read more
2. What advancements in entropy encoding and quantization contribute to improved compression efficiency in High Efficiency Video Coding (HEVC) for lossless and near-lossless video?
This theme investigates refinements in entropy coding and quantization matrices specifically within the HEVC standard, aimed at increasing coding efficiency while managing computational complexity. These contributions address coding of transform coefficients, adoption of optimized weighting parameters, and quantization matrix adaptations to strike an effective balance between bitrate, distortion, and processing speed, crucial for real-time high-resolution video applications and transition towards lossless or near-lossless video compression.
Key finding: The paper presents an enhanced entropy encoding scheme for HEVC by optimizing weighting parameters using the firefly algorithm, demonstrating improved compression rates over the standard entropy coding while maintaining high... Read more
Key finding: This research integrates entropy encoding improvements with an optimized quantization matrix (QM) design in HEVC, demonstrating a compression performance improvement of up to 35.29% over standard encoding schemes. The... Read more
Key finding: This paper proposes a novel context-modeling scheme tailored for coding quantized transform coefficients in larger block sizes (beyond 8x8) within HEVC codecs. Compared to a straightforward CABAC extension, the adaptive scan... Read more
3. How can distributed video coding approaches reduce encoder complexity while achieving low complexity, lossless or low-bit-rate video compression?
This area explores distributed video coding (DVC) frameworks which relocate computational load from encoders to decoders by leveraging principles such as the Slepian-Wolf and Wyner-Ziv theories. Approaches include transform-domain and pixel-domain residual coding, with recent enhancements focusing on decision modes to limit channel coding loads. These techniques are particularly promising for low-power devices or bandwidth-constrained applications requiring low-complexity video encoding without quality loss.
Key finding: The paper introduces QUAM (QUantized TrAnsform Decision Mode), a mechanism that omits zero quantized transform blocks from channel encoding in Distributed Residual Video Coding (DRVC). By reducing the number of bit planes for... Read more