How to write a scientific paper (By Michael LaRocca)
pompous(simple)? circumlocution(single)? irrelevant?
simplicity and clarity
bold 膽大
scientific writers are rarely literate
critical technique關(guān)鍵技術(shù) ?? hand-crafted heurisitics??? sub-optimal次優(yōu) ?? rule-based
leverage杠桿作用(use (something) to maximum advantage) ?? outperforms性能優(yōu)于??? redundancy冗余??? conventionally傳統(tǒng).?? time-consuming耗時(shí)的??? evaluate評估?? infeasible不可行.???
fine-grained細(xì)粒度.?? sensitive敏感.?? trail and error反復(fù)試驗(yàn). ?
penalizing accuracy loss while encouraging model shrinking and speedup懲罰準(zhǔn)確性損失衬以,同時(shí)鼓勵(lì)模型縮小和加速
facilitating促進(jìn).?? validation驗(yàn)證.?? scenarios場景. ? fine-tuning微調(diào). ? protocol協(xié)議. ?
expert-tuned精調(diào).?? factorization因式分解.?? extensive廣泛. ?
pedestrian detection行人檢測. ? state-of-the-art最先進(jìn)的. ? outside the scope of超出范圍. ?
complementary補(bǔ)充. ? distinctive獨(dú)特. ??comprehensive全面.?? versatile多才多藝.??
diminish減少.?? have been borne out已經(jīng)被證實(shí)了??? indispensable必不可少.?? capitalize利用.??
availability可用性.?? immense巨大.?? modality形態(tài).??
yielding further societal,economic, and scientific impacts產(chǎn)生進(jìn)一步的社會功舀,經(jīng)濟(jì)和科學(xué)影響
based on yet unseen data基于尚未看到的數(shù)據(jù). ?
tackle even more challenging problems解決更具挑戰(zhàn)性的問題 ??
perceptual task感知任務(wù).?? the bulk of the paper論文的大部分.?? assignment分配.??
contradictory矛盾??? coordinate協(xié)調(diào)/坐標(biāo).?? consise簡潔.?? signify表示/意味著/象征
unambiguously明白地.??
this latter fact put the second goal at odds with the first后一個(gè)事實(shí)使第二個(gè)目標(biāo)與第一個(gè)目標(biāo)不一致
Sychronization同步.?? heterogeneity異質(zhì)性.?? cognitive認(rèn)知.?? intuitive security直觀安全
Heterogeneous computing異構(gòu)計(jì)算.??
The building blocks of most of scientific computing基石.??