ImageHD delivers up to 40.4x speedup and 383x energy efficiency for on-device continual learning of visual representations by using hyperdimensional computing and bounded exemplar management on an FPGA.
Quantization and training of neural networks for efficient integer-arithmetic-only inference
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
Replacement Learning replaces selected blocks in CNNs and ViTs with learnable parameter-fusion surrogates derived from adjacent layers to reduce full-depth backpropagation redundancy.
citing papers explorer
-
ImageHD: Energy-Efficient On-Device Continual Learning of Visual Representations via Hyperdimensional Computing
ImageHD delivers up to 40.4x speedup and 383x energy efficiency for on-device continual learning of visual representations by using hyperdimensional computing and bounded exemplar management on an FPGA.
-
Replacement Learning: Training Neural Networks with Fewer Parameters
Replacement Learning replaces selected blocks in CNNs and ViTs with learnable parameter-fusion surrogates derived from adjacent layers to reduce full-depth backpropagation redundancy.