ACRC Online Workshop Series on Case Solving

We are delighted to provide you with a recap of our recent online workshop series on “Case Solving,” conducted by case expert Zsolt Ábrahám.

The workshop series consisted of three sessions, which aimed to equip participants with the concepts and tools necessary to excel in case and problem-solving, in addition to other skills relevant to business competitions.

The first session, “Option Tree & the Case Solving Process,” held on January 19th, provided insights into how to set up an option tree – a common tool used in business – and how to analyze the case-solving process effectively.

The second session, “Analysis and Strategy Building for Cases,” held on January 26th, focused on the various tools and techniques used in analyzing business cases. Zsolt shared his experience and insights on how to develop a winning strategy for a case, including understanding key performance indicators (KPIs) and how to work with them.

The final session, “Storylining & Case Presentation,” held on February 2nd, taught attendees about the importance of storytelling and how to use it in business competitions. Zsolt shared his experience on how to create a compelling story through storylining techniques and deliver effective presentations during case competitions.

Stay tuned for updates on our upcoming events!

Other Events
港大經管學院最新多模態AI圖像生成能力排名出爐 部份中國人工智能模型表現突出
2025 | 學院新聞
港大經管學院最新多模態AI圖像生成能力排名出爐 部份中國人工智能模型表現突出
港大經管學院今日發表《人工智能模型圖像生成能力綜合評測報告》,針對15個「文生圖模型」及7個「多模態大語言模型」進行全面評估。研究顯示,字節跳動的即夢AI和豆包,以及百度的文心一言,在新圖像生成的內容質素及圖像修改的表現突出;而早前引起全球關注的DeepSeek最新推出的文生圖模型Janus-Pro,則在新圖像生成方面表現欠佳。研究亦發現部分文生圖模型雖然在內容質素方面表現優異,卻在安全與責任方面的表現強差人意。整體而言,與文生圖模型相比,多模態大語言模型整體表現較佳。
人工智能圖像生成評測成績單公布: 字節跳動百度表現亮眼,DeepSeek Janus-Pro表現欠佳
2025 | 研究
人工智能圖像生成評測成績單公布: 字節跳動百度表現亮眼,DeepSeek Janus-Pro表現欠佳
如今,人工智能領域的前沿模型技術已經從文本處理拓展至視覺信息的深度理解與生成。這些模型既能精准解讀圖像語義,又能根據文字描述創作出兼具真實感與藝術性的視覺內容,展現出令人驚嘆的跨模態理解與創作能力。本研究聚焦全新圖像的生成和基于現有圖像的圖像修改兩大核心任務,提出了一套系統性的人工智能模型圖像生成能力評測框架。我們基于多維測試集的構建與專家評審,對15個專業文生圖模型和7個多模態大語言模型的圖像生成能力進行了全面評估。結果顯示,字節跳動的即夢AI和豆包以及百度的文心一言在新圖像生成的內容質量與修改任務中表現突出,位列第一梯隊。對比不同類型的AI模型,我們發現,相對于專業文生圖模型,多模態大語言模型整體表現更佳。